Linear Polyethyleneimine Grafted to a Hyperbranched Poly(ethylene

Dec 22, 2005 - A block copolymer of a hyperbranched poly(ethylene glycol)-like core and linear polyethylenimine (HBP) was synthesized by a facile synt...
0 downloads 11 Views 257KB Size
Download PDF
Bioconjugate Chem. 2006, 17, 125−131

125

Linear Polyethyleneimine Grafted to a Hyperbranched Poly(ethylene glycol)-like Core: A Copolymer for Gene Delivery Pallab Banerjee,* Ralph Weissleder, and Alexei Bogdanov, Jr.* Center for Molecular Imaging Research, Massachusetts General Hospital, Building 149, 13th Street, Charlestown, Massachusetts 02129. Received March 18, 2005; Revised Manuscript Received November 21, 2005

A block copolymer of a hyperbranched poly(ethylene glycol)-like core and linear polyethylenimine (HBP) was synthesized by a facile synthetic route that included (1) a single-step cationic copolymerization of diepoxy and polyhydroxyl monomers, (2) derivatization of hydroxyl groups of the core HBPEG copolymer with either tosyl or chloromethylbenzoyl chlorides resulting in a corresponding macroinitiator, and (3) synthesis of HBPEG-blockpoly(alkyl oxazolines). HBPEG-block-linear polyethyleneimine (HBP) was obtained by hydrolysis of HBPEGblock-poly(alkyl oxazolines). Linear PEI-bearing hyperbranched polycations (HBP) had lower inherent toxicity in cell culture than PEG-grafted linear polyethyleneimines (PEGLPEI). PEGLPEI formed a complex with DNA with an average diameter of 250 nm. The complexes were loosely condensed and formed aggregates and precipitates during storage. By contrast, hyperbranched polycations (HBP) formed ∼50 nm nanocomplexes with DNA that were stable for several weeks and showed resistance to DNAse I-mediated degradation. The ‘inverted’ block copolymers showed several orders of magnitude higher transfection efficiency than PEGLPEI in vitro. Because of the biocompatibility and higher transfection efficiency, the ‘inverted’ block copolymer merits further investigation as a gene carrier.

INTRODUCTION Polyionic complexes between polycations (e.g. branched polyethyleneimine (BPEI) (1), “fractured” polyamidoamine (PAMAM) dendrimer fragments (2, 3)), and plasmid DNA were shown to transfect cells in vitro with high efficiency. Unfortunately, positively charged complexes with DNA induce acute inflammatory toxicity in vivo associated with the release of TNFR, complement activation, and erythrocyte aggregation (4, 5). Covalent modification of BPEI and PAMAM with poly(ethylene glycol) (2, 6, 7) for providing nonionic shells has been previously used to address the above problem and resulted in an overall lower systemic toxicity (8-11). Since poly(ethylene glycol) (PEG) partially shields a polycation’s positive charge (8-11), large amounts of PEG-grafted polymers are usually required to obtain complexes with DNA for efficient cell transfection (2). Recently, Petersen et al. (12) reported synthesis and DNA-condensing properties of a block copolymer of low molecular weight branched polyethyleneimine grafted to a starpoly(ethylene glycol) core. The grafting of such low molecular mass PEI by block copolymerization enhanced the transfection efficacy of PEI which otherwise shows poor or low transfection (12). An alternative strategy of polycationic hydrophilic corefirst polymer synthesis (13) has been also reported but the efficacy of resultant polymers for DNA delivery is yet to be explored. Our earlier work suggested that although PEG-grafted poly-L-lysine (protected graft copolymer, PGC) (14) showed a long blood circulation time in vivo, low cytotoxicity, and * Correspondence should be addressed to S2-804, Department of Radiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655, Tel. (508) 856-5571, Fax (508) 856-1860. E-mail: [email protected]. 1 Abbreviations: PEG, poly(ethylene glycol); LPEI, linear polyethyleneimine; BPEI, branched polyethyleneimine; HBPEG, hyperbranched poly(ethylene glycol); HBPEOX, hyperbranched poly(ethylene glycol) block poly(ethyl oxazoline); HBPMOX, hyperbranched poly(ethylene glycol) block poly(methyl oxazoline); HBP, hyperbranched poly(ethylene glycol) block linear polyethyleneime.

immunogenicity, its efficacy for in vivo gene transfer was limited (15). One potential explanation of this is in that grafted PEG chains provide shielding of the positive charge of the core polycation (10) and thus usually has limited use for in vivo gene delivery due to low binding to cells (16). However, the lack of efficacy in vitro does not preclude the potential use of PEG-grafted polycations for targeted delivery of expression vectors in vivo. To facilitate higher structural diversity of the “inverted”, core-first polycationic gene delivery carriers, we extended the above studies by devising a hyperbranched PEGlike core polymer (HBPEG) and using the core as a macroinitiator for grafting it with linear polyethyleneimine obtained by using cationic polymerization of 2-alkyl-2-oxazolines. Here we report a comparative study of two classes of polymers: (1) graft copolymers of PEI and PEG and (2) the above novel polycationic block copolymer-hyperbranched polycations (HBP). We compared their ability to form complexes with plasmid DNA, thus providing protection of DNA from nuclease degradation and cell transfection.

MATERIALS AND METHODS Materials. Branched polyethyleneimine (BPEI) (Mn 10 kDa), linear polyethyleneimine (LPEI) (22 kDa) were obtained from Polyscience (Warrington, PA). Other chemicals were from Aldrich or Fisher Scientific. Detailed polymer syntheses are described in Supporting Information. Characterization. Gel permeation chromatography (GPC) was performed by using PL aquagel-OH 40 8mm column (Polymer Laboratories Inc, Amherst, MA) coupled with two detectors, UV/vis, and differential refractive index (RI) at 30 °C. The column was eluted with 100 mM ammonium acetate (pH 7). Molecular mass was measured using poly(ethylene oxide) standards (5-920 kDa Polysciences, Inc, Warrington, PA). The weight average molecular mass was computed using refractive index data. Proton and 13C NMR spectra were recorded in solutions (Spectral Data Service, Champaign, IL) at 400 MHz for 1H and 100 MHz for 13C spectra, respectively.

10.1021/bc050083e CCC: $33.50 © 2006 American Chemical Society Published on Web 12/22/2005

126 Bioconjugate Chem., Vol. 17, No. 1, 2006

The quantitative 13C spectra were acquired in DMSO-d6 using inverse gated decoupling pulse sequences. The spectra were obtained by using 0.5 g/mL polymer solutions at 30 °C. The data were acquired with an 80° pulse and 10 s recycle delay. Cell Culture. COS-1 cells were propagated in DMEM supplemented with 4 mM L-glutamine, 1% penicillin-streptomycin, and 10% fetal bovine serum at 37 °C in 5% CO2. Reporter Gene Plasmid. pCMV-Luc (3.2 MDa, 5.12 kb) plasmid vector encoding firefly luciferase was propagated in E. coli DH5R and using Megaprep kit (Qiagen, Valencia, CA). Plasmid integrity was confirmed by gel electrophoresis in agarose. DNA concentration and purity were determined by measuring absorbance at 260 nm and 280 nm. Polymer-DNA Complexes. To prepare complexes of each individual polymer with plasmid DNA, polymer solutions were added to DNA solutions (20 µg DNA/mL in 25 mM HEPES/ 5% glucose buffer, pH 5.5), mixed by vortexing, and incubated for 20 min at room temperature before use, unless stated otherwise. Polymer stock solutions were diluted with the same buffer as above to desired concentration. Hydrodynamic diameters of complexes were measured daily for four weeks by using Zetasizer 1000HSA (Malvern Instruments Inc, Southborogh, MA), at a fixed angle of 90° at room temperature. All complexes were stored at 4 °C. An ethidium bromide exclusion experiment was performed with a F-4500 fluorescence spectrophotometer (Hitachi, Danbury, CT) at λex ) 366 nm and λem ) 591 nm, respectively. Briefly, 10 µg of DNA and 2 µL of ethidium bromide solution (0.4 mg/mL) were mixed in 2 mL of TE buffer (pH 7.1). Polymer solution was added in 4-5 µL aliquots (1.1 mg/mL).The mixture was incubated at room temperature for 6 min, and the fluorescence intensity was measured. Two consecutive measurements were performed, and intensity values were averaged and corrected for dilution. Cytotoxicity Assay. Cells were plated at a density 10000 cells/well of 96-well plates in 150 mL of medium 24 h before the experiment. Medium was replaced by fresh 150 µL of serumcontaining media before the addition of polymers. Polymer solutions in 50 µL were then added to each well and incubated at 48 h, 37 °C. Cytotoxicity assays (n ) 4 experiments) were performed using CytoTox 96 assay (Promega, Madison, WI), as described by the manufacturer. Nontreated cells were used as controls. Cell Transfection. Transfections with polymer-plasmid DNA complexes were performed, and the results were analyzed using standard assay for luciferase activity expression. Various amounts of polymer in 5% glucose, 25 mM HEPES, were added to the plasmid solution (0.5 µg pCMV-luc plasmid in 30 µL of 0.1 TE buffer), mixed, and incubated at room temperature for 30 min. Cells were plated at a density of 10000 cells/well in 96-well plates, 24 h prior to transfection. Before the addition of complexes, medium was replaced by fresh 100 µL of serum containing media. The complexes were then added to each well. The total volume did not exceed 150 µL/ well. The transfection was carried out for 24 h at 37 °C. The luciferase activity in cell lysates was determined using Microlumat LB96P (EG&G Berthold), by adding 50 µL of luciferin (Promega, Madison, WI) to 25 µL of cell lysate in each well using an automated injector (n ) 3 independent experiments for each transfection reagent). Luciferase activity was integrated over 10 s with 2 s intervals. The protein concentration was measured in the rest of 25 µL cell lysate using standard protein assay reagent PierceEndogen, Rockford, IL). Results were expressed in relative light units per mg of cell protein. The significance of differences between data sets (here and below) was determined by using Student’s unpaired, two-tailed test with Welch’s correction.

Banerjee et al.

DNAse I Protection Assay. DNA was prelabeled with Cy3 by using Label IT Cy3 kit (Mirus, Madison, WI) as described by the manufacturer. Labeled DNA was mixed with unlabeled pCMV-Luc at 1:6 w/w. The final concentration of DNA was 1 µg/µL. HBP and Cy3-labeled pCMV-Luc complexes were prepared at the best transfection efficiency ratio using 1.8 µg of DNA in 10 mM Tris buffer (pH 7.5). The DNA/polymer complex was first incubated for 20 min at room temperature. Various amounts of DNAse I were then added to either “naked” pCMV-Luc or to polymer complexes, and the samples were incubated at room temperature for 150 min. Samples were analyzed by 0.8% agarose gel electrophoresis. Gels were imaged using 530 nm excitation filter (Kodak 440CF) followed by staining with ethidium bromide.

RESULTS Graft Copolymer Synthesis. To prepare a linear PEI backbone polymer (LPEI, Supporting Information, Figure S1A) commercially available poly(2-ethyl 2-oxazoline) was hydrolyzed as described (17). Proton NMR results suggested that hydrolysis was 90% complete. Two graft copolymers were prepared by grafting PEG chains onto linear polyethyleneimine (LPEI) as well as onto the branched polyethyleneimine (BPEI) (Supporting Information, Figure S1). The linear polyethyleneimine graft MPEG (PEGLPEI) was obtained by reacting the MPEG chloroformate with LPEI (Supporting Information, Figure S1A). The graft copolymer contained 5% PEG by weight (i.e. 2-3 PEG chains per molecule of LPEI), as calculated from 1H NMR using the ratio of peak intensity of PEG (-CH O-, 2 3.57 ppm) to ethyleneimine peak intensity (NHCH2CH2, 3.03.3 ppm). Branched PEI (BPEI, Mn 10 kDa) was reacted with monomethoxy PEG carboxy-N-hydroxysuccinimide. The corresponding graft-copolymer obtained using BPEI as a backbone contained approximately 10 wt % of PEG calculated from 1H NMR using the ratio of peak intensity of PEG (-CH2O, 3.57 ppm) to ethyleneimine proton peak intensity (NHCH2CH2, 2.93.2 ppm). Synthesis of Hyperbranched PEG-like Core Polymers (HBPEG). Cationic copolymerization of poly(ethylene glycol) diglycidyl ether (bearing two reactive epoxy groups) and pentaerithrytol ethoxylate (containing four hydroxyl groups) was accomplished by using boron trifluoride as a catalyst (Supporting Information, Figure S2). By varying the catalyst concentration, polymers with different molecular masses ranging from 25 to 35 kDa were obtained. Upon the omission of the polyol (pentaerithrytol ethoxylate), the formation of a gel instead of a soluble polymer was observed. The use of methylene chloride instead of THF as copolymerization solvent resulted in polymers with a lower molecular mass. Quantitative 13C NMR showed incorporation of butylene ether moiety within the HBPEG composition (Figure 1B). The butylene ether moiety was absent from 13C NMR spectra if the copolymerization was carried out in methylene chloride. The hyperbranched PEG-like polymers were isolated as semisolids that had excellent solubility in water and various organic solvents (e.g. methylene chloride, tetrahydrofuran). The obtained HBPEG polymers were analyzed by using a size-exclusion chromatography (Figure 1A, Table 1). Due to the fact that size exclusion chromatography (SEC) tends to underestimate the true molecular mass of branched polymers (18), the true molecular masses of branched polymers were likely to be higher than those suggested by SEC if linear polymers were used as standards. The polydispersity of obtained products was low due to the removal of low molecular weight polymer fraction during the dialysis. The presence of hydroxyl groups was determined by titration method as described (19). Titration of available hydroxyls in HBPEG gave a value of approximately 100 hydroxyl groups/HBPEG molecule. The

A Copolymer for Gene Delivery

Figure 1. A. Gel permeation chromatography of HBPEG2 (Mn 24.7 kDa); “inverted” precursor, a block copolymer HBPMOX1 (Mn 56.2 kDa), and the hydrolyzed product (HBP1) (Mn 42.6kDa). B. Structural units of hyperbranched poly(ethylene glycol)-like core (HBPEG2, Table 1) predicted from quantitative 13C NMR data in DMSO-d6. Linear monomeric unit (l) 44.3 and 72.5-78.2 ppm; branch (t) peaks at 30.4, 44.5, 60.4-60.8, and 62.9 ppm; dendritic core (d) peaks at 29.4 and 72.4 ppm. C. 1H NMR spectra of HBPMOX1 in D2O. Inserts: enlarged view of spectras in the following ranges (1) 1.2-2.2 ppm; (2) 2.6-4.2 ppm; and (3) 7.9-8.2 ppm.

observed high content of terminal hydroxyl groups argues for the formation of extensive branching structure. To calculate the degree of branching, quantitative 13C NMR data were used. We assigned the 44.3, 72.5-78.2 ppm peaks to the linear monomeric unit (l), whereas peaks corresponding to 30.4, 44.5, 60.4-60.8, and 62.9 ppm were assigned to the termini of branches (t). The peaks at 29.4 and 72.4 ppm correspond to the dendritic core (d). The degree of branching was calculated using the following formula as described in (20). The degree of branching, DB ) (Nd + Nt)/(Nd + Nt + Nl), where N is the integrated value of

Bioconjugate Chem., Vol. 17, No. 1, 2006 127

the peak; d, l, t are dendritic, linear, and terminal units, respectively (Figure 1B). The calculated DB values were 0.68 (HBPEG1) and 0.59 (HBPEG2). Both polymers were synthesized using identical conditions and various amounts of the catalyst (Table 1). According to published data, in hyperbranched polymers the degree of branching is approximately 0.5 (21). HBP Block Copolymer Synthesis. We synthesized a cationic block copolymer (HBP) with grafted polyethyleneimine chains using hyperbranched PEG (HBPEG) as the inner core. First, we performed activation of surface hydroxyl groups of HBPEG using either 4-chloromethylbenzoyl chloride or p-toluenesulfonyl chloride. The above chlorides catalyze ring-opening polymerization of 2-oxazoline groups (22, 23) (Supporting Information, Figure S3). The purified chloromethyl or tosyl derivative of HBPEG was characterized by 13C NMR data that showed the presence of a sulfonyl derivative (δ 127.9 (aromatic ortho carbon (CH)2-C-SO2-) and chloromethyl derivative (δ 128.4 (aromatic ortho carbon, ClCH2C(CH)2), respectively. The block copolymer prepared by reacting ethyl oxazoline with the tosyl derivative of HBPEG (HBPEOX) was precipitated in ethyl ether, dried in a vacuum, and characterized (Table 1). 1H NMR peaks at 3.2-3.44 ppm ((-CH2CH2NCOCH2CH2-) and 1.49 ppm (CH3CH2CO-) confirmed the presence of poly(2-ethyl 2-oxazoline). The polymer was then hydrolyzed as described earlier (23), using sodium hydroxide in ethanol/ethylene glycol and purified by dialysis. The disappearance of peak at 3.2-3.4 ppm and the new peak at 2.59-3.09 ppm (-NHCH2CH2-) suggested that hydrolysis of the block copolymer proceeded efficiently and that the “inverted” block copolymer (HBP4) was formed (Table 1). The block copolymer of the chloromethyl derivative of HBPEG was prepared by the cationic polymerization method, using 2-methyl 2-oxazoline monomer in benzonitrile and potassium iodide as a coinitiator. Potassium iodide was used for exchanging chloride anion with iodide, that resulted in high efficacy of conversion into HBPEG-poly(methyl oxazoline) (Supporting Information, HBPMOX, Figure S3). Upon varying the reaction composition and time of polymerization we prepared different molecular weight HBPMOX (Table 1). We further analyzed the obtained polymers using GPC (Figure 1A) before and after the alkaline hydrolysis that demonstrated the progressive increase of the average mass of the polymer after the extension of poly(2-ethyl 2-oxazoline) chains and their hydrolysis (Table 1). Figure 1C depicts the 1H NMR spectrum of HBPMOX1. The observed peak at 7.95 ppm (aromatic ortho hydrogen, -NCH2(CH)2(CH)2-) pointed to the extension of the poly(2-methyl 2-oxazoline) block from the chloromethyl derivative HBPEG core as the reaction scheme (Supporting Information, Figure S3) suggested. The methyl protons of the acetyl group (1.9 ppm) were undetectable after the hydrolysis, whereas methylene protons adjacent to nitrogen shifted from 3.4 to 2.7 ppm. Complex Formation with DNA. The formation of complexes between DNA and HBP polymers was monitored by fluorometry using the ethidium bromide displacement method. Control PEG-grafted polymers (PEGLPEI and PEGBPEI) showed a gradual decrease of fluorescence that reached 50% of the initial relative fluorescence intensity value in the presence of large polymer amounts (concentration exceeded 50 µg/mL, Figure 2). In the case of hyperbranched block copolymer (HBP), linear polyethyleneimine LPEI, and commercially available branched polyethyleneimine BPEI (Figure 2), a 50% decrease of fluorescence intensity was observed in the presence of much lower polymer amounts (approximately 10-13 µg/mL). DNA condensing properties of polymers were tested in parallel by measuring hydrodynamic radii of polymers and the complex with negatively charged DNA (Table 1). Initially, we measured

128 Bioconjugate Chem., Vol. 17, No. 1, 2006

Banerjee et al.

Table 1. A Characterization of Hyperbranched and Graft Copolymers and Their Precursors polymer HBPEG1c HBPEG2d HBPEOX HBPMOX1 HBPMOX2e HBPMOX3f HBP1g HBP2h HBP3i HBP4j PEGLPEI PEGBPEI

Mn (K)

MW (K)

PDI (Mw/Mn)

yield (%)

HDa (nm)

CHDb (nm)

CHD4Wk (nm)

Zeta Plm (mV)

38.1 24.7 67.3 56.2 49.3 43.6 42.6 37.5 29.3 49.3 36.8 20.4

72.1 41.3 120 97.5 81.5 69.5 69.4 59.6 48.9 75.9 78.0 45.5

1.89 1.67 1.78 1.73 1.65 1.59 1.63 1.59 1.67 1.54 2.12 2.23

92 90 44 82 90 89 -

30.9 ( 3.9 24.7 ( 5.6 48.7 ( 4.8 39.8 ( 3.9 32.8 ( 6.3 27.8 ( 5.7 46.1 ( 5.8 31.2 ( 3.4 17.2 ( 2.7 60.9 ( 6.2 273.4 ( 5.8 256.7 ( 12.3

35.9 ( 5.3 56.2 ( 8.5 56.1 ( 6.7 44.6 ( 3.4 195 ( 10.9 73.4 ( 8.7

36.9 ( 3.7 54.2 ( 4.3 61.3 ( 6.2 43.7 ( 4.1

11.72 4.69 -

a Hydrodynamic diameter; measured at 50 µg polymer/mL. b Hydrodynamic diameter of DNA complex prepared by mixing 25 µg of plasmid DNA and 50 µg of polymer in 1 mL 5% glucose solution. c 0.01 mol boron trifluoride etherate. d 0.02 mol boron trifluoride etherate. e 10 mL of 2-methyl 2-oxazoline, polymerization time 24 h. f 7 mL 2-methyl 2-oxazoline, polymerization time 24 h. g Synthesized from hydrolyzed HBPMOX1. h Synthesized from hydrolyzed HBPMOX2. i Synthesized from hydrolyzed HBPMOX3. j Synthesized from hydrolyzed HPPEOX. k Hydrodynamic diameter of the DNA complex after 4 weeks of storage. l Size exclusion chromatography was performed in 0.1 M ammonium acetate using poly(ethylene oxide)s as molecular weight standards. m Complex was prepared by mixing 25 µg of DNA and 50 µg of polymer in 1 mL of 5% glucose solution.

Figure 2. DNA-bound ethidium bromide displacement as a result of complex formation with polycations. The legend denotes complexes formed with BPEI (closed circles), LPEI (closed triangles), PEGBPEI (open circles), PEGLPEI (diamonds), and HBP1 (squares). The y-axis represents the normalized fluorescence intensity. Results of three independent experiments are shown as mean ( SD.

the hydrodynamic diameter of the HBPEG core (approximately 25 nm). After block copolymerization with the formation of the LPEI “corona”, the diameter increased by ∼10 nm. Further, we observed that complexes between the plasmid DNA and HBP formed in nonionic, isotonic solution (5% glucose in water) were small, ∼50 nm in diameter, and that particle size did not change for at least 30 days. We further performed measurements of zeta potentials of complexes prepared using 25 µg of DNA and 50 µg of hyperbranched block copolymer (HBP2 or HBP3). Zeta potential of complexes was affected by changing the density of cationic residues in the polymer. We used two different molecular weight polymers for zeta potential measurement and established that higher molecular weight polymer (HBP2) yields more positive zeta potential of the complex than the low molar mass polymer HBP3 (Table 1). The ability of the HBP to protect the nucleic acids from enzymatic cleavage was tested by using DNAse I treatment, and the data were compared to that obtained using LPEI and the conventional graft copolymer (PEGLPEI). The complex composition corresponded to that with the highest transfection efficiency in cell culture (see below). HBPs were able to protect DNA from degradation by DNAse I for at least 2 h (Figure 3), and only traces of DNA fragments were present at the highest

concentrations of DNAse I used. The control “naked” DNA as well as PEGLPEI complex with DNA was completely degraded at concentrations of DNAse I exceeding 2.5 U/mL. In the case of the complex prepared using LPEI, ethidium bromide staining of the complex is much higher than the in the complexes prepared using HBPs. Cell Transfection. Cell transfection efficacy was compared in COS-1 cell culture. Prepared in identical conditions, all complexes gave very different levels of luciferase expression. In case of both types of “conventional” graft copolymers transfection efficiency was low even when higher amounts of polymer were used for complex preparation (Figure 4A). The transfection efficacy observed in the case of HBP block copolymers was compared against linear polyethylneimine (LPEI). The resultant luciferase activity expressed in transfected cells was at least 5 orders of magnitude higher than any of PEGLPEI or PEGBPEI while the amount of hyperbranched block copolymer needed for transfection was less (at least 50 times). Used at similar weight ratio, hyperbranched block copolymers showed at least two orders higher transfection efficiency than that of positive control Superfect reagent (Figure 4A and Figure 4B). At slightly higher concentrations, hyperbranched block copolymer showed higher transfection efficiency than LPEI. Cytotoxicity. Cytotoxicity of polymers was investigated in COS-1 cells at 48 h after the addition of polymers (Figure 4C). The highest molecular mass HBP had a very low cytoxicity even at a very high concentration (0.5 mg/mL). Although at lower concentration the toxicity of HBP was lower than that of LPEI, the observed differences were not significant.

DISCUSSION It is well-known that upon mixing positively charged block copolymers with DNA, polycationic blocks induce DNA binding and condensation while PEG blocks provide steric protection for condensed DNA-polycation cargo (7, 24). The major goal of our research was to compare “conventional” copolymers composed of polycationic PEI- and PEG-based blocks with HBPs, based on hyperbranched PEG-like polymer block surrounded with PEI blocks, i.e. “inverted” design. The “conventional” copolymers contained between 5 and 10% of PEG2000 by weight, respectively, which appears to prevent binding of the DNA-polymer complex to the cell (16, 25), resulting in at least 4 orders of magnitude lower transfection efficiency than that of its 22 kDa PEG-free LPEI counterpart at all concentrations (26). Our experiments with similar copolymers also showed that even at very high concentrations of graft copolymers (1.5 mg/mL), the highest achieved transfection efficiency

A Copolymer for Gene Delivery

Bioconjugate Chem., Vol. 17, No. 1, 2006 129

Figure 3. DNAse I protection assay. Lane 1 shows control pCMV-Luc/Cy3-pCMV-Luc mixture (6:1 by weight, 1.5 µg of DNA was loaded per lane). Lane 2-7 shows the same mixture at different amounts of DNAse I present: 2.5U, 5U, 7.5U, 10U, 50U, 100U, and 250U, respectively. DNAse I was incubated with DNA and polycation-containing samples for 2.5 h at room temperature followed by 0.8% agarose gel electrophoresis and imaging using Kodak 440CF equipped with Cy3 fluorescence filters (upper row), staining with ethidium bromide, and imaging (lower row). A typical result is shown.

Figure 4. A. Luciferase expression in COS 1 cells after a 24 h transfection in serum-containing medium: transfection efficiency of PEGLPEI (solid bars), PEGBPEI (open bars) at various N/P ratios. Transfection results obtained using various concentrations of Superfect reagent are shown in the inset. Polymer complex with plasmid DNA was prepared by mixing 0.5 µg of pCMV-luc in 30 µL solvent with polymers at different concentrations. B. Luciferase expression in COS-1 cells after a 24 h transfection in serum-containing medium: transfection efficiency of LPEI (solid bars), HBP3 (shaded bars), and HBP2 (open bars). The complex was prepared by mixing 0.5 µg of pCMV-luc plasmid DNA with various polymer amounts (shown in x-axis). C. Cytotoxicity profiles of linear polyethyleneimine (closed squares) and hyperbranched block copolymers: HBP3 (closed triangles), HBP4 (open squares), PEGLPEI (open circles), and PEGBPEI (open triangles) in COS-1 cell culture after 48 h incubation. Results of four independent experiments are shown as mean ( SD.

was still at least 3 orders of magnitude lower than that of LPEI (Figure 4A,B). Ethidium bromide displacement results suggested that for the purposes of cell transfection, “conventional” PEGgrafted polymers are effectively shielded by PEG, resulting in

only partial availability of positive charges of the polymer for complex formation with the DNA (Figure 2). In choosing HBP architecture, we hypothesized that the PEGbased polymer core could still provide protection to bound DNA

130 Bioconjugate Chem., Vol. 17, No. 1, 2006

while the polycationic “shell” could be fully functional in DNA binding. Thus, we expected that upon the formation of complexes with negatively charged DNA, the cationic residues would get neutralized and that the formation of nanosized complexes could be achieved. We devised the synthesis of the above hyperbranched molecule using a mixture of poly(ethylene glycol) diglycidyl ether and pentaerythritol ethoxylate upon cationic polymerization. The method enabled us to obtain high molecular mass hydroxyl-terminated, hyperbranched PEG-like polymer (HBPEG) in a single step. The available hydroxyl groups in hyperbranched polymer could be easily modified using chloromethyl benzoyl chloride or p-toluene sulfonyl chloride, yielding a hyperbranched macroinitiator that enabled synthesis of a block copolymer with poly(2-alkyl 2-oxazoline). The molar mass of block copolymer (Mn) could be controlled by changing the reaction parameters (Table 1). By treating polyoxazoline blocks in mild alkaline conditions, we obtained final HBPEGLPEI (HBP). Unlike PEGylated PEI (e.g. PEGLPEI or PEGBPEI), HBPmediated condensation of DNA appeared not to be affected by hyperbranched PEG-like core blocks, because the concentration of HBP required to achieve a half-maximal displacement of ethidium bromide initially bound to DNA in the complex was low (Figure 2). Consequently, the DNA condensing properties of the “inverted” block copolymer was superior to that of graft copolymers used in this study. At the same time, particle size and stability data suggested that HBP behaved differently from PEGBPEI, since the complex with plasmid DNA was very stable during the storage and was at least two times smaller than DNA complexes obtained using “conventional” PEG-grafted PEI (Table 1, see also ref 24). The observed lack of core interference with the complex formation was also apparent in experiments which investigated DNA-protective properties of HBP. We observed that ethidium bromide-negative, i.e., highly condensed complexes, were highly resistant to DNAse I degradation (Figure 3). We further measured cytotoxicity as well as transfection efficacy of the resultant block copolymer and compared them with the PEG-grafted PEI and LPEI (22 kDa). Due to the presence of hyperbranched HBPEG core, a higher amount of block copolymer was required to achieve optimal cell transfection (Figure 4A) at the concentration that was less toxic than that of LPEI (Figure 4C). It is known that PEG-grafted BPEI always show less toxicity than BPEI alone (10). In our case, i.e., when PEG-like polymer comprised the inner core of the molecule, the effect of lesser toxicity was achieved in a manner similar to that of the “conventional” PEG coating (Figure 4C). This suggests that the presence of HBPEG inner core indeed reduced cytotoxicity, whereas complex condensation by “inverted” block copolymer HBPs was superior to that of PEG grafted PEI (26). The latter effect could be explained by the lack of interference with DNA binding to PEI. In conclusion, we report here synthesis and properties of hyperbranched block copolymers based on PEG-like hyperbranched core-first design. We demonstrated that “inverted” block copolymers (HBPs) obtained using grafting of linear PEI showed low toxicity in vitro and formed stable nanosized complexes with plasmid vectors that were protected from degradation with DNAse I. Furthermore, these nanocomplexes showed high transfection efficacy in vitro. Since in vitro transfection efficacy is not ideal in predicting the ability of a polycation to deliver expression or silencing vectors in vivo, the latter properties of hyperbranched polycations with PEGlike cores are currently under investigation.

Banerjee et al.

ACKNOWLEDGMENT This work has been supported in part by 1P50CA86355-01 (Project 2). Supporting Information Available: Experimental details. This material is available free of charge via the Internet at http:// pubs.acs.org.

LITERATURE CITED (1) Boussif, O., Lezoualch, F., Zanta, M. A., Mergny, M. D., Scherman, D., Demeneix, B., and Behr, J. P. (1995) A versatile vector for gene and oligonucleotide transfer into cells in culture and in-vivo polyethylenimine. Proc. Natl. Acad. Sci. U.S.A. 92, 7297-7301. (2) Luo, D., Haverstick, K., Belcheva, N., Han, E., and Saltzman, W. M. (2002) Poly(ethylene glycol)-conjugated pamam dendrimer for biocompatible, high-efficiency DNA delivery. Macromolecules 35, 3456-3462. (3) Tang, M. X., Redemann, C. T., and Szoka, F. C. (1996) In vitro gene delivery by degraded polyamidoamine dendrimers. Bioconjugate Chem. 7, 703-714. (4) Ogris, M., Brunner, S., Schuller, S., Kircheis, R., and Wagner, E. (1999) Pegylated DNA/transferrin-pei complexes: Reduced interaction with blood components, extended circulation in blood and potential for systemic gene delivery. Gene Ther. 6, 595-605. (5) Plank, C., Mechtler, K., Szoka, F. C., Jr., and Wagner, E. (1996) Activation of the complement system by synthetic DNA complexes: A potential barrier for intravenous gene delivery. Hum. Gene Ther. 7, 1437-46. (6) 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. (7) Bronich, T. K., Cherry, T., Vinogradov, S. V., Eisenberg, A., Kabanov, V. A., and Kabanov, A. V. (1998) Self-assembly in mixtures of poly(ethylene oxide)-graft- poly(ethyleneimine) and alkyl sulfates. Langmuir 14, 6101-6106. (8) Kabanov, A. V., Vinogradov, S. V., Suzdaltseva, Y. G., and Alakhov, V. Y. (1995) Water-soluble block polycations as carriers for oligonucleotide delivery. Bioconjugate Chem. 6, 639-643. (9) Katayose, S., and Kataoka, K. (1997) Water-soluble polyion complex associates of DNA and poly(ethylene glycol)-poly(l-lysine) block copolymer. Bioconjugate Chem. 8, 702-707. (10) Petersen, H., Fechner, P. M., Fischer, D., and Kissel, T. (2002) Synthesis, characterization, and biocompatibility of polyethylenimine-graft-poly(ethylene glycol) block copolymers. Macromolecules 35, 6867-6874. (11) Kircheis, R., Wightman, L., and Wagner, E. (2001) Design and gene delivery activity of modified polyethylenimines. AdV. Drug DeliVery ReV. 53, 341-358. (12) Petersen, H., Kunath, K., Martin, A., Stolnik, S., Roberts, C., Davies, M., and Kissel, K. (2002) Star-shaped poly(ethylene glycol)block-polyethylenimine copolymers enhance DNA condensation of low molecular weight polyethylenimines. Macromolecules 3, 926936. (13) Koc¸ , F., Wyszogrodzka, M., Eilbracht, P., and Haag, R. (2005) Highly regioselective synthesis of amino-functionalized dendritic polyglycerols by a one-pot hydroformylation/reductive amination sequence. J. Org. Chem. 70, 2021-2025. (14) Bogdanov, A. A., Weissleder, R., Frank, H. W., Bogdanova, A. V., Nossif, N., Schaffer, B. K., Tsai, E., Papisov, M. I., and Brady, T. J. (1993) A new macromolecule as a contrast agent for mr-angiography - preparation, properties, and animal studies. Radiology 187, 701-706. (15) Bogdanov, A., Kayne, L., and Weissleder, R. (1998) in The 25th International Symposium on Controlled Release of BioactiVe Materials (Controlled Release Society, Intl., Ed.) pp 91-92, Controlled Release Society, Inc., Las Vegas, NV. (16) Nguyen, H. K., Lemieux, P., Vinogradov, S. V., Gebhart, C. L., Guerin, N., Paradis, G., Bronich, T. K., Alakhov, V. Y., and Kabanov, A. V. (2000) Evaluation of polyether-polyethyleneimine graft copolymers as gene transfer agents. Gene Ther. 7, 126-138.

Bioconjugate Chem., Vol. 17, No. 1, 2006 131

A Copolymer for Gene Delivery (17) 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-2-oxazoline). J. Controlled Release 73, 391-399. (18) Feast, W. J., and Stainton, N. M. (1995) Synthesis, structure and properties of some hyperbranched polyesters. J. Mater. Chem. 5, 405-411. (19) Sunder, A., Turk, H., Haag, R., and Frey, H. (2000) Copolymers of glycidol and glycidyl ethers: Design of branched polyether polyols by combination of latent cyclic ab(2) and abr monomers. Macromolecules 33, 7682-7692. (20) Hawker, C. J., Lee, R., and Frechet, J. M. J. (1991) One-step synthesis of hyperbranched dendritic polyesters. J. Am. Chem. Soc. 113, 4583-4588. (21) Radke, W., Litvinenko, G., and Muller, A. H. E. (1998) Effect of core-forming molecules on molecular weight distribution and degree of branching in the synthesis of hyperbranched polymers. Macromolecules 31, 239-248. (22) Kobayashi, S. (1990) Ethylenimine polymers. Prog. Polym. Sci. 15, 751-823.

(23) Akiyama, Y., Harada, A., Nagasaki, Y., and Kataoka, K. (2000) Synthesis of poly(ethylene glycol)-block-poly(ethylenimine) possessing an acetal group at the peg end. Macromolecules 33, 58415845. (24) Petersen, H., Fechner, P. M., Martin, A. L., Kunath, K., Stolnik, S., Roberts, C. J., Fischer, D., Davies, M. C., and Kissel, T. (2002) Polyethylenimine-graft-poly(ethylene glycol) copolymers: Influence of copolymer block structure on DNA complexation and biological activities as gene delivery system. Bioconjugate Chem. 13, 845854. (25) Erbacher, P., Bettinger, T., Belguise-Valladier, P., Zou, S. M., Coll, J. L., Behr, J. P., and Remy, J. S. (1999) Transfection and physical properties of various saccharide, poly(ethylene glycol), and antibody-derivatized polyethylenimines (pei). J. Gene. Med. 1, 210222. (26) Kichler, A., Chillon, M., Leborgne, C., Danos, O., and Frisch, B. (2002) Intranasal gene delivery with a polyethylenimine-peg conjugate. J. Controlled Release 81, 379-388. BC050083E