Polyplexes Assembled with Internally Quaternized PAMAM-OH

Nov 4, 2003 - Polyplexes Assembled with Internally Quaternized PAMAM-OH Dendrimer and Plasmid DNA Have a Neutral Surface and Gene Delivery ...
0 downloads 0 Views 188KB Size
Download PDF
1214

Bioconjugate Chem. 2003, 14, 1214−1221

Polyplexes Assembled with Internally Quaternized PAMAM-OH Dendrimer and Plasmid DNA Have a Neutral Surface and Gene Delivery Potency Jung Hoon Lee, Yong-beom Lim, Joon Sig Choi, Yan Lee, Tae-il Kim, Hyun Jin Kim, Jae Keun Yoon, Kwan Kim, and Jong-sang Park* School of Chemistry & Molecular Engineering, Seoul National University, Kwanak-ku, Seoul 151-742, Korea. Received June 5, 2003; Revised Manuscript Received September 11, 2003

Interior tertiary amine groups of PAMAM-OH dendrimers (hydroxyl-terminated polyamidoamine, PAMAM) were modified by methylation to make these polymers have a more cationic character, which enabled electrostatic interaction between PAMAM-OH and plasmid DNA. A methylation reaction was dose-dependent, producing internally quaternized PAMAM-OH (QPAMAM-OH), thereby making tertiary amine/quaternary amine ratio adjustment possible. More highly condensed particles of plasmid DNA were formed as the degree of quaternization increased, whereas unmodified polymer (PAMAMOH) could not. The location of positive charges in the internal position of QPAMAM-OH resulted in the formation of neutral polyplexes in which ζ potential leveled off near the zero value even at high charge ratios (+/-) of 10. A light scattering experiment showed that the polyplex formed by QPAMAMOH was very small with the size of 53.3 nm at the optimum condition. QPAMAM-OH/DNA polyplexes were round-shaped with the more compact and small particles formed as the charge ratio increased. QPAMAM-OH showed much reduced cytotoxicity compared with starburst PAMAM and branched polyethyleneimine (PEI) in which shielding of interior positive charges by surface hydroxyls might be the reason for this favorable result. These results suggest that QPAMAM-OH could be a promising tool as a nonviral vector both by itself and in conjugated form with targeting ligands.

INTRODUCTION

Numerous gene delivery systems based on viral (1-3) and nonviral (4, 5) vectors have been developed and tried so far. Recently, several recurring issues about safety of viral vectors have led to a careful reconsideration of the use of them in human clinical trials such as in Gelsinger’s case (6, 7). Moreover, they have significant limitation in large-scale production and available DNA size they can carry. In response to these problems, nonviral gene delivery systems such as cationic polymers or cationic lipids have attracted great attention to achieve a breakthrough in the development of an “ideal” gene carrier (8). Several polymeric materials have been investigated as candidates for gene delivery; among them, cationic polymers with hydrophilic segments are gaining attention (9-14). It is because hydrophilic segments detoxify cationic polymers (11, 12), improve solubility (12, 13, 15) and prevent polycation/DNA complexes from aggregation in vivo. Starburst PAMAM dendrimers are highly branched spherical polymers characterized by primary amine groups at the surface and tertiary amine groups in the interior. The polyplexes formed by PAMAM showed efficient transfection in which the primary amine groups participate in DNA binding and the tertiary amine groups exert endosome buffering effect (14, 16, 17). Biological behavior of dendrimers depends to a large extent on their surface groups. In one example, it was * To whom correspondence should be addressed. Jong-sang Park, School of Chemistry & Molecular Engineering, Seoul National University, San 56-1, Shillim-dong, Kwanak-ku, Seoul 151-742, Korea, Tel: 82 -2 -880-6660, Fax: 82-2-877-5110, E-mail: [email protected].

reported that amine-terminated PAMAM dendrimer interacted with bovine serum albumin more strongly than carboxyl-terminated one (18). Serum albumin has been considered to cause aggregation of nonviral carriers in vivo (19). PAMAM-OH dendrimers are structurally identical to PAMAM except that surface amine functions have been replaced by hydroxyl groups. Absence of surface primary amine groups in PAMAM-OH makes this polymer nearly neutral which might be advantageous in terms of cytotoxicity and an aggregation problem. However, PAMAMOH is nearly unable to form DNA polyplex because of the low pKa of interior tertiary amines (20). To overcome this hurdle, we introduced internal quaternary ammonium salt to the tertiary amine of PAMAMOH dendrimers by methylation in order to provide binding sites for negatively charged plasmid DNA. Some polymers containing quaternary ammonium groups have been known to make polyplexes with DNA more efficiently than the polymers having primary or secondary amine groups. In addition, the polyplexes of quaternary amine-based polymers were relatively smaller compared to those prepared by cationic polymers containing primary and tertiary amine groups (21). Quaternary aminebased polymers are cationic at most pHs as a strong polyelectrolyte, while the charge density in the primary amine-based polymers depend on the pH of media (22). From such a viewpoint, here we report the synthesis and characterization of internally quaternized PAMAMOH. We expected that interior quaternary amine groups of QPAMAM-OH would interact negatively charged DNA while preserving a neutral polymer and/or a polyplex surface, which would act affirmatively with regard to cytotoxicity and an aggregation behavior of polyplexes.

10.1021/bc034095g CCC: $25.00 © 2003 American Chemical Society Published on Web 11/04/2003

PAMAM-OH Dendrimer Containing Internal Quaternary Amine Table 1. Quaternization of PAMAM-OH G4a PAMAM-OH

0.27 Q

0.52 Q

0.78 Q

0.97 Q

CH3I temperature

0.5 equiv 25 °C

0.75 equiv 37.5 °C

0.9 equiv 37.5 °C

4.16 equiv 37.5 °C

a Equiv means the equivalent moles of methyl iodide relative to interior tertiary amines in PAMAM-OH G4.

EXPERIMENTAL PROCEDURES

Materials. PAMAM-OH G4, PAMAM G4 (Starburst), Methyl iodide, anhydrous N,N-dimethlyformamide (DMF), and PEI (average molecular weight 25 kDa) were purchased from Aldrich (Milwaukee, WI). PGL3-control vector (plasmid DNA) was purchased from Promega (Madision, WI). Fetal bovine serum (FBS) and Dubecco’s modified Eagle’s medium (DMEM) were purchased from GIBCO (Gaithersburg, MD). Synthesis of Quaternized PAMAM-OH (QPAMAMOH). The solvent (methanol) was vacuum-evaporated and dried from manufacturer’s PAMAM-OH solution prior to reaction. After redissolving PAMAM-OH (0.1 g, 7 µmol) in DMF (0.5 mL), methyl iodide of various molar ratios diluted in DMF (0.5 mL) was added. The mixture was stirred at each optimal reaction temperature (Table 1). After 24 h, the mixture was precipitated into diethyl ether and vacuum-dried and the residue obtained was redissolved in 1 mL of water. The solution was placed into a dialysis membrane (SpectraPore, MwCO 60008000) and dialyzed against 2 M NaCl and pure water in succession. Freeze-drying of water resulted in a white powder of QPAMAM-OH. 1H NMR (300 MHz, D2O) δ 2.47 (br m, CH2CO), 2.69 (br m, NCH2CH2NHCO), 2.88 (br m, NCH2CH2), 3.15 (s, CH3), 3.34 (br m, CH2CH2OH), 3.52 (br m, CH2N+), 3.66 (br m, CH2OH). Ethidium Bromide Exclusion Assay. Ethidium bromide (1.0 µg) in 10 µL of water and plasmid DNA (1.0 µg) in 10 µL of water were mixed for 10 min at room temperature. After incubation, the plasmid DNA/ethidium bromide mixture was added to quaternized PAMAMOH dendrimers with various charge ratios, ranging from 0.25 to 10 (+/-) and incubated further for 30 min. The charge ratio was calculated by relating the number of quaternary amine groups of QPAMAM-OH derivatives and the number of phosphate groups of DNA. The complexes were diluted to a total of 2 mL of Hepes buffered saline (HBS, 25 mM Hepes, 150 mM NaCl, pH 7.4) prior to measuring fluorescence intensity with a spectrofluorometer (JASCO FP-750). Excitation (λex) and emission (λem) wavelengths were 260 and 600 nm, respectively. The fluorescence of the DNA solution in HBS with ethidium bromide was calculated as 100%. The buffer containing ethidium bromide only without DNA was used as a blank control. Dynamic Light Scattering (DLS) Measurement. The size of complexes was determined using a BI-200SM Goniometer (Brookhaven Instruments Corporation, Holtsvile, NY) with a Lexel laser model 95 argon laser (100 mW output power at a wavelengh of 514.5 nm). Correlator, PD2000 (Precision Detectors) was used and the scattering angle was 90°. Complexes were formed at a final concentration of 5 µg/mL plasmid DNA in water. DNA stock solution was added to QPAMAM-OH derivatives or PAMAM G4 prepared at various concentrations. DLS was performed in triplicate with the sampling time set to automatic. ζ Potential Measurement. The complexes of PAMAM or 0.97 QPAMAM-OH and plasmid DNA at various charge ratios were prepared in water. Surface charges were measured using a Zetasizer (Malvern Instrument

Bioconjugate Chem., Vol. 14, No. 6, 2003 1215

Ltd, Malvern UK) equipped with a He-Ne laser at a wavelength of 680 nm. Atomic Force Microscopy (AFM). Atomic force microscopy (Nanoscope IIIa System, Digital Instruments, Inc., Santa Barbara, CA) was used for imaging the shape of complexes at 4 (N/P or ( ratio). Complexes were formed at a total of 1 µg/mL of plasmid DNA concentration in water. In the case of PAMAM-OH/plasmid mixture, 2.5 mM of MgCl2 was applied to the solution. Complexes containing 1 ng of plasmid DNA were applied to freshly cleaved mica and incubated on the mica for 5 min. After incubation, excess fluid was wicked off using filter paper. The solution was dried at room temperature prior to imaging. The image mode was set to tapping mode and average scan speed was 2 Hz. Cytotoxicity Assay in Vitro. For the cytotoxicity assay, an MTT assay was performed (23). 293T cells were seeded at a density of 1 × 104 cells/well in a 96-well plate and grown in 95 µL of DMEM containing 10% FBS for 24 h, supplying 5% CO2 at 37 °C. After treating cells with PEI 25 kDa, PAMAM G4, PAMAM-OH G4, and QPAMAM-OHs for 1 day, 25 µL of MTT stock solution (5 mg/ mL) was added to each well and incubated for 2 h. Then, 100 µL of extraction buffer (20% w/v of SDS in 50% DMF, pH 4.7) was added. Absorbance was measured at 570 nm after overnight incubation. Transfection. 293T cells (5 × 104 cells/well) were seeded in 24-well plates and grown in 600 µL of DMEM containing 10% FBS for 1 day. Polyplexes of plasmid DNA and dendrimers were prepared by mixing 0.5 mL of plasmid DNA (4 µg/mL) and 0.5 mL of PAMAM or QPAMAM-OHs at various N/P or charge ratios, respectively, in FBS-free DMEM, and the mixtures were incubated for 30 min at room temperature. Following 4 h treatment of polyplexes, the medium was replaced by 1 mL of DMEM containing 10% FBS. Cells were incubated further for 2 days at 37 °C. After the growth medium was removed, cells were washed with PBS and lysed for 30 min at room temperature by 100 µL of Reporter lysis buffer (Promega, Madision, WI). The lysate was cleared by centrifugation. Luciferase activity was measured using a LB 9507 luminometer (Berthold, Germany), and the protein content was measured by Micro BCA assay reagents (Pierce, Rockford, IL). RESULTS AND DISCUSSION

Synthesis of QPAMAM-OHs. QPAMAM-OHs with various degrees of internal quaternization were synthesized by partial or near complete methylation of interior tertiary amine groups (Scheme 1). The peak area of N+methyl group was compared to the total proton numbers of unmodified PAMAM-OH, as confirmed by empirical estimation and theoretical calculation by 1H NMR (Figure 1). From the 1H NMR spectra, the degree of quaternization was determined by dividing the experimental proton numbers of the N-methyl groups (3.15 ppm) with the calculated total proton numbers that were obtained by assuming that all the tertiary amines were completely methylated. Careful control over the reaction conditions, e.g., the amount of methyl iodide and reaction temperature, made it possible to produce QPAMAM-OHs with various degrees of quaternization. 1H NMR spectra showed that surface hydroxyl groups were not reacted with methyl iodide. The significant chemical shift changes occurred at carbons c, d due to the chemical environment of positively charged nitrogen. Chloride salt of QPAMAMOHs with the various quaternization degree of 27% (0.27 QPAMAM-OH), 52% (0.52 QPAMAM-OH), 78% (0.78

1216 Bioconjugate Chem., Vol. 14, No. 6, 2003

Lee et al.

Scheme 1. Synthesis of QPAMAM-OHa

Figure 1. 1H NMR spectra of QPAMAM-OH derivatives. (A) PAMAM-OH G4, (B) 0.27 QPAMAM-OH, (C) 0.52 QPAMAMOH, (D) 0.78 QPAMAM-OH, and (E) 0.97 QPAMAM-OH. c′ and d′ are shifted peaks from c and d after quaternization.

a Tertiary amine groups of PAMAM-OH were converted into quaternary amine groups at a degree of 27 % (0.27 QPAMAMOH), 52 % (0.52 QPAMAM-OH), 78 % (0.78 QPAMAM-OH) and 97 % (0.97 QPAMAM-OH), respectively.

QPAMAM-OH), and 97% (0.97 QPAMAM-OH) were obtained after dialysis and lyophilization. The PAMAM-OH dendrimers were transformed to be functional by equipping quaternary amine groups for DNA condensation, the hydroxyl groups of their surface for low cytotoxicity, and internal positively charged amines by modifying tertiary amines to quaternary ammonium salts, which screened their charges due to the exterior hydroxyl groups leading to neutral surface charges. Analysis of Complex Formation by Agarose Gel Electropheresis and Ethidium Bromide Exclusion Assay. The PAMAM-OHs having various degrees of quaternization and plasmid DNA were mixed, and the mixtures were electrophoresed in agarose gel to see if

polyplexes could be formed by interior positive charges (Figure 2). No indication of polyplex formation between PAMAM-OH and DNA was found (Figure 2A). Low pKa values of PAMAM-OH’s interior tertiary amines and ensuing low charge density should be the reason for this. Slowly migrating bands in Figure 2B might be the polyplex of 0.27 QPAMAM-OH and DNA; however, no complete retardation of DNA was observed in this polymer. It was observed that the number of charges per copolymer was an important factor in condensing DNA into small particles and in determining other physicochemical characteristics of a polymer/DNA complex (24). As the degree of quaternization increased, complete retardation of DNA was observed (Figure 2C, 2D, and 2E). More highly quaternized polymers were more efficient in polyplex formation. The polyplex formation in 0.97 QPAMAM-OH was nearly stoichiometric. As shown in Figure 3, ethidium bromide exclusion assay was performed to quantify and confirm the complex formation ability of QPAMAM-OHs. Initial fluorescence of DNA and ethidium bromide complex decreased as dendrimers bound to DNA, releasing intercalated ethidium bromide. In 0.27 QPAMAM-OH, the relative fluorescence intensity decreased to 70% at a charge ratio of 5 and plateaued after that point. A low charge density of 0.27 QPAMAM-OH might be the reason and it is in line with the electrophoresis result (vide ante). DNA was completely retarded by 0.52 QPAMAM-OH above a charge ratio of 2 in the electrophoresis experiment; however, ethidium exclusion did not go beyond around 60% even at a charge ratio of 10. The unexpected phenomenon was revealed that interaction between 0.52 QPAMAM-OH and DNA is too weak to change DNA conformation sufficiently to exclude intercalated ethidium. 0.78 QPAMAM-OH and 0.97 QPAMAM-OH expelled ethidium from DNA, completely above a charge ratio of about 2. In summary, as the degrees of quaternization

PAMAM-OH Dendrimer Containing Internal Quaternary Amine

Bioconjugate Chem., Vol. 14, No. 6, 2003 1217

Figure 2. Agarose gel band shift assay. (A) PAMAM-OH/plasmid DNA polyplexes, (B) 0.27 QPAMAM-OH/plasmid DNA polyplexes, (C) 0.52 QPAMAM-OH/plasmid DNA polyplexes, (D) 0.78 QPAMAM-OH/plasmid DNA polyplexes, and (E) 0.97 QPAMAM-OH/plasmid DNA polyplexes. Charge ratios (+/-) are indicated above each lane. In the case of PAMAM-OH/DNA, numbers of each lane are represented N/P ratios. The charge ratio was calculated by relating the number of quaternary amine groups of QPAMAM-OH derivatives and the number of phosphate groups of DNA, and the N/P ratio was calculated from the number of tertiary amines of PAMAM-OH and the number of phosphate groups of DNA. Samples were electrophoresed in 0.7% agarose gel at 100 V for 40 min in Tris-borate buffer.

Figure 4. Surface charge of particles measured by ζ potential experiments. The polyplexes of 0.97 QPAMAM-OH ([) or PAMAM G4 (0) and plasmid DNA were formed at 0.5, 1, 4, and 10 of charge ratios (+/-). Results are shown as mean ( standard deviation (n ) 3).

Figure 3. Ethidium bromide exclusion assay. N/P ratio of PAMAM-OH G4 and charge ratio (+/-) of QPAMAM-OH/ plasmid DNA was 0.25, 0.5, 0.75, 1, 2, 5, and 10. Data are expressed as a mean relative fluorescence intensity (%, n ) 3) at each ratio and the mean ( standard deviation are shown at each data point. Table 2. The Complex Size of QPAMAM-OH/DNA Polyplexes and PAMAM/DNA Polyplexes As Determined by Dynamic Light Scattering (+/-)

PAMAM

0.52 Q

0.78 Q

0.97 Q

1 2 4 10

168.0 126.0 86.4 78.0

680.2 122.6 90.9 66.1

157.0 108.4 84.8 51.3

129.3 111.3 74.0 53.3

a Data are the mean diameter (nm) of each polyplex observed in water at 5 µg/mL of plasmid DNA concentration. Charge ratio in PAMAM was calculated on the assumption that only primary amines are protonated in near neutral pH

(97% > 78% > 52% > 27%) in QPAMAM-OH increased, the formation of polyplexes was more efficient. Characterization of QPAMAM-OHs/DNA Particles. For analysis of particle size, we performed dynamic light scattering measurements. Briefly, in the presence of quaternary amine-based PAMAM-OH, relatively smaller, compared to primary amine-based PAMAM, particles were formed (Table 2). Interestingly, in the case of 0.52 QPAMAM-OH, very large particles (680.2

nm) were detected at a charge ratio of 1 due to the poor ability of complex formation with DNA. This is also seen from the neutralization between DNA and 0.52 QPAMAMOH, in which cancelation of net charge (25) and minimization of charge-to-charge repulsion between complex particles might cause the large particles (26). However, increasing the charge ratio up to 2, the particle size became small (122.6 nm). The size distribution of the particles of more quaternized QPAMAM-OHs with DNA was below 157 nm at all charge ratios, which means these particles were small enough to be taken up by receptormediated endocytosis requiring a smaller size than 150 nm (27). It can be also noted that the diameter of complex particles gradually decreased as the degree of quaternization in QPAMAM-OH increased. This result indicates that the more compact polyplexes are produced at higher degrees of QPAMAM-OHs. We believe from these results that plasmid DNA having a extended structure can contact and bind to internal positive charges. The surface charges of particles were determined by ζ potential measurements (Figure 4). The surface charge of PAMAM G4 particles increased as increasing charge ratios and positive ζ potentials were observed above a charge ratio of 1. For QPAMAM-OH, although the charge ratio was increased up to 10, its surface charge never went to values beyond zero. Starburst PAMAM-OH dendrimers are highly branched macromolecules that have a specific size, spherical shape, and rigid conformation, and it seems likely that the dendrimer has enough

1218 Bioconjugate Chem., Vol. 14, No. 6, 2003

Lee et al.

Figure 5. Particle morphology imaged by AFM at charge ratio (+/-) ) 4. (A) PAMAM G4, (B) PAMAM-OH G4, (C) 0.27 QPAMAMOH, (D) 0.52 QPAMAM-OH, (E) 0.78- QPAMAM-OH, and (F) 0.97 QPAMAM-OH.

inner space (28) to accommodate large plasmid DNA molecules. This should be the reason hydroxyl groups exposed at their surface are able to screen internal positive charges even increasing their charge ratios. Thus, the condensates are considered to be electrostatistically neutralized at their surface. The morphology and size of the complexes depending on the degree of quaternization was investigated by

atomic force microscopy (AFM) at a charge ratio of 4.0 (Figure 5). Only free plasmid DNA, or very large and loose condensates were observed for PAMAM-OH/DNA polyplexes and 0.27 QPAMAM-OH/DNA polyplexes. 0.52 QPAMAM-OH showed partially condensed structures of DNA. Theses results were reconfirmed by ethidium bromide exclusion assay results in which no sharp decrease of relative fluorescence was observed by 0.52

PAMAM-OH Dendrimer Containing Internal Quaternary Amine

Bioconjugate Chem., Vol. 14, No. 6, 2003 1219

Figure 6. Cytotoxicity assay on 293T cell after exposing the cells with PAMAM-OH, QPAMAM-OH, PAMAM G4, and PEI 25kDa at various concentrations. Relative cell viability (RCV, %) is expressed as a percentage of viable cells divided by untreated cells.

Figure 7. Transfection efficiency in 293T cell at charge ratio (+/-) ) 6. Data are expressed as a RLU (Relative light unit) per µg protein.

QPAMAM-OH. The images of polyplexes composed of 0.78 QPAMAM-OH and 0.97 QPAMAM-OH were observed to be small and spherical particles. Cytotoxicity Issue. Cytotoxic effects of polycations are mainly mediated by interactions of the polymers with cell membranes and/or by cellular uptake of the polymers, and subsequent activation of intracellular signal transduction pathways (29). We examined the cytotoxicity of QPAMAM-OHs with various degrees of quaternization, comparing it with PAMAM G4 and branched PEI (25 kDa) (Figure 6). The cells exposed to PAMAM G4 were more viable than that treated with PEI. In general, the relative cell viability (RCV) of both PEI and PAMAM G4 decreased significantly as the concentraion increased. However, quaternized PAMAM-OHs were less toxic than PAMAM G4 and PEI 25kDa. The RCV (%) of all quaternized PAMAM-OH derivatives was over 90% throughout all the concentration levels tested. The cytotoxicity of cationic macromolecules has been studied and reported extensively so far. One hypothesis is that their toxicity results from the interaction between positively charged polymers and negatively charged cell membranes leading to a haemolytic effect (30, 31). The presence of primary amines has a significant toxic effect on red blood cells, causing them to agglutinate (32). So, many strategies to decrease cytotoxicity of polymers were reported as follows. The cytotoxicity of polyallyamine was reported to be decreased by substituting primary amine groups with amido hydroxyl groups (11). It was also reported that the quaternization of the PEI polymer resulted in a decrease of toxicity (33). The fact that all QPAMAM-OHs were less cytotoxic than their parent dendrimer (PAMAM) should be the result of shielding interior positive charges by surface hydroxyls and of quaternization. Dendrimer-shielding of a surface charge confirmed by ζ potential measurement might be the major reason for reduced toxicity, minimizing the direct contact of these polycations to cell membrane. In addition, an inverse correlation between cytotoxicity and degree of quaternization was observed: PAMAM-OH > 0.26 QPAMAM-OH > 0.52 QPAMAM-OH > 0.78 QPAMAMOH ≈ 0.97 QPAMAM-OH with cell cytotoxicity mirroring

the degree of quaternization. The results suggest that the cytotoxicity is a function of the nature of the polycation charged moiety (primary, secondary, tertiary, or quaternary amine group) rather than its charge density (34). The quaternary amines were always charged and strongly hydrophilic; thus, the majority of quaternary amine-containing PAMAM-OHs appeared to be less toxic than tertiary amine-containing PAMAM-OH. Moreover, the higher density of quaternary amine-containing PAMAM-OH benefited from improved hydrophilicity and showed less cytotoxicity. Therefore, the transformation of the water-soluble polymers had previously resulted in a marked reduction in toxicity (33, 35). Introducing quaternary amines to PAMAM-OH resulted not only in less cytotoxicity, but also the ability for condensation with plasmid DNA while the parent PAMAM-OH could not. These hydroxyl group-modified dendrimer-based delivery systems are supposed to be useful for reduction of the toxicity of PAMAM dendrimers. Transfection Efficiency in Vitro. The gene delivery efficiency of QPAMAM-OH/DNA polyplexes as judged by luciferase gene expression was tested on 293T cell lines. The result (Figure 7) indicated that the transfection efficiency of QPAMAM-OH was lower than that of PAMAM G4 and PEI. The neutral surface charge of QPAMAM-OH/DNA polyplexes, as shown by measurement of the ζ potential, might decrease chances of the polyplexes to bind electrostatistically to cell-matrix and cell-cell anchoring proteins, such as heparan sulfate proteoglycans (36, 37). In principle, cellular uptake of a polycation is a nonspecific process. A positively charged complex binds to a negatively charged cell membrane with electrostatic interaction. However, anionic proteoglycans are too ubiquitous for cationic particles to reach specific targeting organs in vivo. Therefore, the introduction of a system which decreases the surface charge of particles is necessary to avoid such a nonspecific uptake into cells. QPAMAM-OHs have such a low nonspecific interaction potential due to the neutral surface of particles; however,

1220 Bioconjugate Chem., Vol. 14, No. 6, 2003

they should be equipped with cell-surface targeting residues that could trigger their receptor-mediated endocytosis. Although the transfection efficiency of QPAMAM-OH derivatives was lower by 1 order of magnitude than PAMAM G4, our particles have the merit of much lower cytotoxicity and the chance to attach a ligand at the hydroxyl groups of their surface. These properties are useful for the strategy of receptor-mediated transfection by attaching a ligand as is being done in our further works. CONCLUSION

We introduced quaternary ammonium salts that were placed in an internal location of starburst PAMAM-OH dendrimer (generation 4) by methylation of interior tertiary amines of PAMAM-OH. Improvements were made, not only in the ability to bind to DNA, but also in the cytotoxicity. The surface hydroxyl functionalities of QPAMAM-OHs could provide low cytotoxicity, and the polyplex of the polymer exhibited a neutral surface charge. Another major advantage of QPAMAM-OH dendrimers is that unreacted hydroxyl groups can be exploited for the conjugation of targeted ligands in order to provide cell entry by receptor-mediated endocytosis, while maintaining decreased nonspecific interaction with ubiquitous cellular proteins. ACKNOWLEDGMENT

We acknowledge the support of this work by the Korea Research Foundation (2001-015-DP0344), the Korea Science and Engineering Foundation (R02-2002-00000011-0), and the SRL-Molecular Therapy Research Center, in Sungkyunkwan University. We thank Prof. Doo Soo Chung for giving access to DLS. LITERATURE CITED (1) Verma, I. M., and Somia, N. (1997) Gene therapy - promises, problems and prospects. Nature 389 239-242. (2) Lotze, M. T., and. Kost, T. A. (2002) Viruses as gene delivery vectors: application to gene function, target validation, and assay development. Cancer Gene Ther. 9, 692-699. (3) Walther, W., and Stein, U. (2000) Viral vectors for gene transfer: a review of their use in the treatment of human diseases. Drugs 60, 249-271. (4) Nishikawa, M., and Huang, L. (2001) Nonviral vectors in the new millennium: delivery barriers in gene transfer. Hum. Gene Ther. 12, 861-870. (5) Li, S., and Huang, L. (2000) Nonviral gene therapy: promises and challenges. Gene Ther. 7, 31-34. (6) Ferber, D. (2001) Gene therapy. Safer and virus-free? Science 294, 1638-1642. (7) Lehrman, S. (1999) Virus treatment questioned after gene therapy death. Nature 401, 517-518. (8) Brown, M. D., Schatzlein, A. G., and Uchegbu, I. F. (2001) Gene delivery with synthetic (non viral) carriers. Int. J. Pharm. 229, 1-21. (9) Choi, J. S., Lee, E. J., Choi, Y. H., Jeong, Y. J., and Park, J. S. (1999) Poly(ethylene glycol)-block-poly(l-lysine) Dendrimer: Novel linear polymer/dendrimer block copolymer forming a spherical water-soluble polyionic complex with DNA. Bioconjugate Chem. 10, 62-65. (10) Kim, T. I., Jang, H. S., Joo, D. K., Choi, J. S., and Park, J. S. (2003) Synthesis of diblock copolymer, methoxypoly(ethylene golycol)-block-polyamidoamine dendrimer and its generation-depenent self-assembly with plasmid DNA. Bull. Korean Chem. Soc. 24, 123-125. (11) Boussif, O., Delair, T., Brua, C., Veron, L., Pavirani, A., and Kolbe, H. V. (1999) Synthesis of polyallylamine deriva-

Lee et al. tives and their use as gene transfer vectors in vitro. Bioconjugate Chem. 10, 877-883. (12) Ahn, C. H., Chae, S. Y., Bae, Y. H., and Kim, S. W. (2002) Biodegradable poly(ethylenimine) for plasmid DNA delivery. J. Controlled Release 80, 273-282. (13) Ward, C. M., Pechar, M., Oupicky, D., Ulbrich, K., and Seymour, L. W. (2002) Modification of pLL/DNA complexes with a multivalent hydrophilic polymer permits folate-mediated targeting in vitro and prolonged plasma circulation in vivo. J. Gene Med. 4, 536-547. (14) Luo, D., Haverstick, B., Nadya. K., Han, E., and Saltzman, W. M. (2002) Poly(ethylene glycol)-Conjugated PAMAM Dendrimer for Biocompatible, High-Efficiency DNA Delivery. Macromolecules 35, 3456-3462. (15) Mannisto, M., Vanderkerken, S., Toncheva, V., Elomaa, M., Ruponen, M., Schacht, E., and Urtti, A. (2002) Structureactivity relationships of poly(L-lysines): effects of pegylation and molecular shape on physicochemical and biological properties in gene delivery. J. Controlled Release 83, 169182. (16) Kukowska-Latallo, J. F., Bielinska, A. U., Johnson, J., Spindler, R., Tomalia, D. A., and Baker, J. R., Jr. (1996) Efficient transfer of genetic material into mammalian cells using Starburst polyamidoamine dendrimers. Proc. Natl. Acad. Sci. U.S.A. 93, 4897-4902. (17) Eichman, J. D., Bielinska, A. U., Kukowska-Latallo, J. F., and Baker, J. R., Jr. (2000) The use of PAMAM dendrimers in the efficient transfer of genetic material into cells. Pharm. Sci. Teck. Today 3, 232-245. (18) Klajnert, B., and Bryszewska, M., (2002) Fluorescence studies on PAMAM dendrimers interactions with bovine serum albumin. Bioelectrochemistry 55, 33-35. (19) Godbey, W. T., and Mikos, A. G. (2001) Recent progress in gene delivery using nonviral transfer complexes. J. Controlled Release 72, 115-125. (20) Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Kallos, G., Martin, S., Roeck, J., Ryder, J., and Smith, P. (1985) A new class of polymers: Starburst-dendritic macromolecules. Polym. J. 17, 177-182. (21) Wolfert, M. A., Dash, P. Nazarova, R., O., Oupicky, D., Seymour, L. W., Smart, S., Strohalm, J., and Ulbrich, K. (1999) Polyelectrolyte vectors for gene delivery: influence of cationic polymer on biophysical properties of complexes formed with DNA. Bioconjugate Chem. 10, 993-1004. (22) Bronich, T. K., Nguyen, H. K., Eisenberg, A., and Kabanov, A. V., (2000) Recognition of DNA Topology in Reactions between Plasmid DNA and Cationic Copolymers. J. Am. Chem. Soc. 122, 8339-8343. (23) Hansen, M. B., Nielsen, S. E., and Berg, K. (1989) Reexamination and further development of a precise and rapid dye method for measuring cell growth/cell kill. J. Immunol. Methods 119, 203-210. (24) Choi, J. S., Joo, D. K., Kim, C. H., Kim, K., and Park, J. S. (2000) Synthesis of a Barbell-like Triblock Copolymer, Poly(L-lysine) Dendrimer-block-Poly(ethylene glycol)-block-Poly(L-lysine) Dendrimer, and Its Self-Assembly with Plasmid DNA. J. Am. Chem. Soc. 122, 474-480. (25) Yaroslavov, A. A., Sukhishvili, S. A., Obolsky, O. L., Yaroslavova, E. G., Kabanov, A. V., and Kabanov, V. A. (1996) DNA affinity to biological membranes is enhanced due to complexation with hydrophobized polycation. FEBS Lett. 384, 177-180. (26) Jeong, J. H., Song, S. H., Lim, D. W., Lee, H., and Park, T. G. (2001) DNA transfection using linear poly(ethylenimine) prepared by controlled acid hydrolysis of poly(2-ethyl-2oxazoline). J. Controlled Release 73, 391-399. (27) Perales, J. C., Ferkol, T., Molas, M., and Hanson, R. W. (1994) An evaluation of receptor-mediated gene transfer using synthetic DNA-ligand complexes. Eur. J. Biochem. 226, 255266. (28) Carnahan, M. A., and Grinstaff, M. W. (2001) Synthesis and characterization of polyether-ester dendrimers from glycerol and lactic acid. J. Am. Chem. Soc. 123, 2905-2906. (29) Fischer, D., Lib, Y., Ahlemeyerc, B., Krieglsteinc, J., and Kissel, T. (2003) In vitro cytotoxicity testing of polycations:

PAMAM-OH Dendrimer Containing Internal Quaternary Amine influence of polymer structure on cell viability and hemolysis. Biomaterials 24, 1121-1131. (30) Malik, N., Wiwattanapatapee, R., Klopsch, R., Lorenz, K., Frey, H., J. Weener, W., Meijer, E. W., Paulus, W., and Duncan, R. (2000) Dendrimers: relationship between structure and biocompatibility in vitro, and preliminary studies on the biodistribution of 125I-labeled polyamidoamine dendrimers in vivo. J. Controlled Release 65, 133-148. (31) Jevprasesphant, R., Penny, J., Jalal, R., Attwood, A., Mckeown, N. B., and D’Emanuele, A. (2003) The influence of surface modification on the cytotoxicity of PAMAM dendrimers. Int. J. Pharm. 252, 263-266. (32) Dekie, L., Toncheva, V., Dubruel, P., E., Schacht, H. Barrett, L., and Seymour, L. W. (2000) Poly L-glutamic acid derivatives as vectors for gene therapy. J. Controlled Release 65, 187-202. (33) Brownlie, A., Uchegbu, I. F., and Schatzlein, A. G. (2002) Controlled Release Society 29th Annual Meeting Proceedings, #105, Controlled Release Society, Memphis, TN.

Bioconjugate Chem., Vol. 14, No. 6, 2003 1221 (34) Zelikin, A. N., Putnam, D., Shastri, P., Langer, R., and Izumrudov, V. A. (2002) Aliphatic ionenes as gene delivery agents: Elucidation of structure-function relationship through modification of charge density and polymer length. Bioconjugate Chem. 13, 548-553. (35) Brown, M. D., Schatzlein, A., Brownlie, A., Jack, V., Wang, W., Tetley, L., Gray, A. I., and Uchegbu, I. F. (2000) Preliminary characterization of novel amino acid based olymeric vesicles as gene and drug delivery agents. Bioconjugate Chem. 11, 880-891. (36) Blessing, T., Remy, J. S., and Behr, J. P. (1998) Monomolecular collapse of plasmid DNA into stable virus-like particles. Proc. Natl. Acad. Sci. U.S.A. 95, 1427-1431. (37) Zuber, G., Dauty, E., Nothisen, M., Belguise, P., and Behr, J. P. (2001) Towards synthetic viruses. Adv. Drug Delivery Rev. 52, 245-253.

BC034095G