Physicochemical and Biological Characterization of Polyethylenimine

It has been reported that the transfection efficiency of PEG-PEI complexes with pSV-β-gal plasmid was slightly decreased with an increase in PEG cont...
0 downloads 0 Views 193KB Size
Download PDF
Bioconjugate Chem. 2004, 15, 677−684

677

ARTICLES Physicochemical and Biological Characterization of Polyethylenimine-graft-Poly(ethylene glycol) Block Copolymers as a Delivery System for Oligonucleotides and Ribozymes Carola Brus,† Holger Petersen,† Achim Aigner,‡ Frank Czubayko,‡ and Thomas Kissel*,† Department of Pharmaceutics and Biopharmacy, Philipps-University, Ketzerbach 63, 35037 Marburg, Germany, and Department of Pharmacology and Toxicology, School of Medicine, Philipps-University, Karl-von-Frisch-Strasse 1, 35033 Marburg, Germany. Received September 9, 2003; Revised Manuscript Received April 28, 2004

Two different series of polyethylenimine (PEI) block copolymers grafted with linear poly(ethylene glycol) (PEG) were investigated as delivery systems for oligodeoxynucleotides (ODN) and ribozymes. The resulting interpolyelectrolyte complexes were characterized with respect to their physicochemical properties, protection efficiency against enzymatic degradation, complement activation, and biological activity under in vitro conditions. The effect of PEG molecular weight and the graft density of PEG blocks on complex characteristics was studied with two different series of block copolymers. The resulting ODN complexes were characterized by photon correlation spectroscopy (PCS) and laser Doppler anemometry (LDA) to determine complex size and zeta potential. Electrophoresis was performed to study the protective effects of the different block copolymers against enzymatic degradation of ODN. Intact ODN was quantified via densitometric analysis. Ribozymes, a particularly unstable type of oligonucleotides, were used to examine the influence of block copolymer structure on biological activity. The stabilization of ribozymes was also characterized in a cell culture model. Within the first series of block copolymers, the grafted PEG chains (5 kDa) had marginal influence on the complex size. Two grafted PEG chains were sufficient to achieve a neutral zeta potential. Within the second series, size and zeta potential increased with an increasing number of PEG chains. A high number of short PEG chains resulted in a decrease in complex size to values comparable to that of the homopolymer PEI 25 kDa and a neutral zeta potential, indicating a complete shielding of the charges. Complement activation decreased with an increasing number of short PEG 550 Da chains. Ribozyme complexes with PEG-PEI block copolymers achieved a 50% down-regulation of the target mRNA. This effect demonstrated an efficient stabilization and biological activity of the ribozyme, which was comparable to that of PEI 25 kDa. PEGylated PEI block copolymers represent a promising new class of drug delivery systems for ODN and ribozymes with increased biocompatibility and physical stability.

INTRODUCTION

Oligodeoxynucleotides (ODN) and ribozymes have stimulated significant interest as tools for sequence-specific down-regulation of gene expression (1-3). To improve limited cellular uptake and poor stability of ODN, several carrier systems are currently under investigation, e.g., cationic lipids (4) and cationic polymers (5). Recently, important discoveries have led to a deeper understanding of the self-assembly processes between cationic polymers and DNA (6-10). Complexation with polycations by electrostatic interactions causes condensation of the DNA (11). The ratio of DNA to polymer, expressed as the N/P ratio, determines the extent of condensation (12). Furthermore, DNA stability against enzymatic degradation * To whom correspondence should be addressed. Tel: (0049)6421-2825881; fax: (0049)-6421-2827016; e-mail: kissel@ mailer.uni-marburg.de. † Department of Pharmaceutics and Biopharmacy. ‡ Department of Pharmacology and Toxicology.

is increased by complexation with the polycations (13). Although, similar mechanisms have been proposed for polyplexes containing ODN (10, 14), they are considered to be one of the most difficult classes of polynucleotides to be complexed by and delivered with polycations. It has been postulated that the limited number of negative charges presented by ODN may be insufficient for cooperative binding (15). Additionally, such interpolyelectrolyte complexes often suffer from a poor stability in solution leading to rapid aggregation (14, 16, 17). To overcome this problem, block copolymers containing a water-soluble nonionic segment, such as poly(ethylene glycol) (PEG), have been recently introduced. These include PEG-polyspermine (14), PEG-polylysine (PLL) (9, 10), and di- or trimethylammonioethyl methacrylate (DMAEMA or TMAEMA)-based copolymers coupled to PEG or other hydrophilic polymers such as poly[N-(2-hydroxypropyl)methacrylamide] (pHPMA) (18, 19), PLLpHPMA (18), PLL-dextran (20), and PEG-polyethylenimine (PEI) (7). Promising results were reported using

10.1021/bc034160m CCC: $27.50 © 2004 American Chemical Society Published on Web 06/08/2004

678 Bioconjugate Chem., Vol. 15, No. 4, 2004

these water-soluble block copolymers for ODN complexation, indicating an increased solubility and stability against enzymatic degradation (7). Cell culture and in vivo experiments have also demonstrated a decreased cytotoxicity (14) and high antisense activity (21-23). A possible explanation for these positive features of ODN copolymer complexes could be the arrangement of the ODN. It is hypothesized that the ODN is trapped within a polycationic core with the hydrophilic polymer chains oriented to the surrounding aqueous environment, creating a hydrophilic shell (16). This core shell structure possibly prevents aggregation due to steric repulsion, thus stabilizing the particles in solution. Furthermore, enzymatic attacks are prevented by the surrounding hydrophilic shell, leading to increased nuclease resistance (7, 10). The degree of PEG substitution was found to influence physicochemical properties of the resulting complexes such as complex size, nuclease resistance (9), and even complement activation (24), posing restrictions for intravenous application of these delivery systems, as demonstrated by the appearance of serious side effects under in vivo conditions (25). Studies using plasmid DNA further revealed an influence of PEG molecular weight on condensation properties, biocompatibility, and biological activity (26). The aim of this study was to establish the structureproperty relationships of block copolymer ODN and ribozyme complexes. The influence of PEGylation degree and PEG molecular weight on physicochemical and biological properties was systematically investigated. MATERIALS AND METHODS

Materials. The 20 mer DNA oligodeoxynucleotide (ODN) with the sequence 5′ TTC CAA TAC AGA AAC TCT CT- 3′ was purchased from Biospring (Frankfurt, Germany), dissolved in sterile water and stored at -20 °C until use. The 40 mer ribozyme Luc Rz 227 with the sequence 5′-ACU GCA UAC CUG AUG AGU CCG UUA GGA CGA AAC GAU UCU G-3′, directed against the luciferase mRNA, was purchased from GENSET S.A (Paris, France). PEI 25 kDa (Polymin water free, 99%) was a gift from BASF (Ludwigshafen, Germany). PEGmonomethyl ether (mPEG) 550 and mPEG 5000 were obtained from Aldrich (Taufkirchen, Germany). Block copolymers were synthesized by grafting linear PEG onto branched PEI using a diisocyanate linker as described in detail previously (27, 28). Polymers were used as sterile filtered aqueous stock solutions (0.9 mg/mL) of pH 7.4. All other materials were of analytical grade. Preparation of Polymer-ODN Complexes. Complexes were prepared as previously published by diluting the appropriate amount of ODN stock solution (100 µg/mL in sterile water), as well as polymer stock solution (0.9 mg/mL in sterile water, pH 7.4) in 50 µL of 150 mM NaCl, vortexing and incubating for 10 min (13). The polymer solution was added to the ODN solution, vortexed again, and incubated for a further 10 min. The polymer/DNA ratio was expressed as the nitrogen/ phosphate (N/P) ratio and calculated on the basis that 1 µg of DNA corresponds to 3 nmol of phosphate, while 1 µL of PEI stock solution contains 10 nmol of nitrogen (13). For photon correlation spectroscopy and transfection experiments complexes were prepared in 150 mM NaCl. For atomic force microscopy, 10 mM NaCl was used to prevent salt crystal formation. Photon Correlation Spectroscopy (PCS). The hydrodynamic diameter of freshly prepared complexes with a ODN concentration of 40 µg/mL were measured using

Brus et al.

a Zetasizer 3000 HS form Malvern Instruments (Herrenberg, Germany) equipped with a 10 mW HeNe laser at a wavelength of 633 nm at 25 °C. Scattered light was detected at 90° angle through a 400 µm pinhole, and the viscosity and refractive index of pure water at 25 °C were used for data analysis. Measurements were analyzed using the CONTIN algorithm. Values given are the means of 10 runs of 60 s ( standard deviation. The instrument was routinely calibrated with Standard Reference Latex Particles (AZ 55 Electrophoresis Standard Kit, Malvern Instruments). Laser Doppler Anemometry (LDA). The zeta potential of freshly prepared complexes with an ODN concentration of 40 µg/mL was determined using the standard capillary electrophoresis cell of the Zetasizer 3000 HS form Malvern Instruments at position 17.0 at 25 °C. Average values were calculated with the data of 10 runs ( standard deviation. Atomic Force Microscopy (AFM). AFM images were conducted on a Dimension3000 Scanning Probe Microscope from Digital Instruments (Santa Barbara, CA) with a Nanoscope IIIa controller. All imaging was carried out in the tapping mode at a scan speed of approximately 2 Hz with 512 × 512 pixel data acquisition. V-shaped cantilevers with a pyramidal tip of silicon nitride were used (Park Scientific, CA). Nuclease Stability. Freshly prepared complexes were incubated with 5 units DNase 1 (Roche Diagnostics GmbH, Mannheim, Germany) per µg DNA for 3 h at 37 °C. The samples were then transferred to a water bath of 70 °C for 30 min to inactivate the nuclease. The ODN was displaced from the complex by adding 8 µL 10% (w/v) sodium dodecyl sulfate (SDS) in water, vortexing, and incubating for a further 10 min. A 10 µL volume of each sample was loaded onto a 15% polyacrylamide gel containing 7 M urea, corresponding to total amount of 0.2 µg of ODN. Electrophoresis was carried out at 200 V for approximately 50 min (Consort E332 power supply, Roth, Karlsruhe, Germany). Afterward, gels were stained for 5 min in a 1:1000 dilution of SYBR Gold (MoBiTech GmbH, Go¨ttingen, Germany) in water and digitalized using the BioDocAnalyze gel documenting system (Biometra, Go¨ttingen, Germany) with an excitation wavelength of 312 nm. Densitometric Analysis. Electrophoretic gels were analyzed using the software program Scion Image 4.0.2. beta (Scion Corporation, Frederick, MD). The amount of intact ODN was calculated via a four point standard curve on every single gel using three ODN dilutions of 0.2, 0.1, and 0.05 µg, respectively, plus the point of origin. Values given are the difference between the calculated ODN amount of each sample and the intact amount of the noncomplexed ODN control and are expressed as the percentage which was protected by the complex. The linear relationship between the amount of the ODN and the optical density of the bands was previously proven by a calibration curve. Five independent calibration curves using nine different dilutions were measured. Linear regression of the means exhibited a correlation coefficient of 0.9947. Differences were considered significant for P e 0.05 based on One-way ANOVA using Origin software package (Microcal Software, Inc. Northampton, MD). Complement Activation. The principle of the hemolytic assay is based on the concentration-dependent consumption of complement components by complementactivating agents when serum is preincubated with these agents (24). The subsequent hemolysis of antibodysensitized sheep erythrocytes by the so-called membrane

Delivery System for Oligonucleotides and Ribozymes

Bioconjugate Chem., Vol. 15, No. 4, 2004 679

Table 1. Nomenclature of the Two Grafted PEI-g-PEG Block Copolymer Series. It Is Further Given the Average PEI Content and the Schematic Structure of the Single Copolymers. The Black Part of the Structure Represents the Branched PEI 25 KDa, and the Gray Part Represents the Nonionic Linear PEG Segments

attack complex formed by the remaining complement components is decreased as compared to that of untreated serum. Gelatin veronal buffer (GBV) and antibodysensitized sheep erythrocytes were both purchased from Sigma-Aldrich (Taufkirchen, Germany). The required serum dilution of 1:10 in GBV, necessary for the lysis of 50% of the erythrocytes, was determined using the Krogh equation as described (29) and expressed as CH50max value. The procedure was performed according to the protocol by Plank et al. (24). Briefly, five different concentrations (25 mg/L; 12.5 mg/L; 6.25 mg/L; 1.56 mg/L and 0.39 mg/L) of freshly prepared bPEI-g-lPEG-ODN complexes at N/P ratio 20 were preincubated with serum. The resulting CH50 of the serum, after preincubation with complexes, was compared to CH50max and expressed as percentage of CH50max ) (CH50 sample/CH50max) × 100 (24). Cell Culture. SKOV-3 ovarian carcinoma cells (American Type Culture Collection, Manassas, VA), stably transfected with the luciferase gene, were cultured in Iscove’s modified Dulbecco’s medium IMDM (PAA Laboratories, Co¨lbe, Germany) supplemented with 10% FCS at 37 °C and 5% CO2 (v/v). Ribozyme Activity Assay. SKOV-3 cells, stably transfected with the luciferase gene (Aigner and UrbanKlein, unpublished results), were seeded at a density of 34 000 (Figure 6a and b) or 70 000 (Figure 6c) cell/well in 24-well plates, respectively. After 6 h, medium was exchanged and the cells were incubated with freshly prepared complexes. Ribozyme-PEI complexes were prepared according to the standard protocol containing 0.25 µg of ribozyme directed against the luciferase mRNA and unrelated ribozyme as a control where indicated. Experiments were carried out at two different N/P ratios of 6.7 and 20 for PEI homopolymers and all PEI-g-PEG block copolymers. Cells incubated with 0.25 µg of noncomplexed ribozyme were used as a control. Medium was exchanged after 4 h. Luciferase activity was determined after a further 44 h using the Luciferase Assay System from Promega according to the manufacturer’s instructions. All experiments were performed in triplicates. The protein concentration in each sample was determined using a BCA assay (30). RESULTS AND DISCUSSION

Two series of block copolymers were chosen to study the influence of PEGylation on the physicochemical properties and biological activity of the resulting inter-

polyelectrolyte complexes with ODN and ribozymes. In both series, branched (b) PEI 25 kDa was grafted with an increasing number of linear (l) PEG chains, with a molecular weight of 5 kDa within the first and 550 Da within the second series, respectively (Table 1). These graft (g) copolymers were designated using the following nomenclature: bPEI(25k)-g-lPEG(x)n with b and l denoting a branched or linear structure, respectively. The number in brackets (25k or x, where x ) 5k or 550) represent the MW of the PEI or the PEG block, and the index n is the average number of PEG blocks per PEI molecule. This number was calculated based on 1H NMR spectra as described previously (27). Photon Correlation Spectroscopy (PCS). The hydrodynamic diameters of ODN-PEG-PEI complexes were determined at two different nitrogen to phosphate (N/P) ratios of 6.7 and 20 (Figure 1). All PEG-PEIs formed very small complexes in the range of 7 to 40 nm. No specific trend could be observed within the first series (Figure 1a). The complex size was only marginally influenced by the increasing degree of PEG substitution with 5 kDa PEG chains. At both N/P ratios, the average diameters were below 30 nm. bPEI(25k)-g-lPEG(5k)15 formed larger complexes of 40 nm at N/P 6.7, possibly due to a decreased condensation efficiency of this highly grafted derivative. At N/P 20, small complex sizees were also observed. A decrease in complex size at higher N/P ratios using PEI 25 kDa has been previously reported for unmodified ODN and a phosphorothioate oligonucleotide previously (31). Petersen et al. also found smaller complex sizes when using five different PEI-g-PEG block copolymers with plasmid DNA at different N/P ratios (26). In the second series, polymer grafted with shorter 550 Da PEG chains, up to a substitution degree of 50, showed increases in complex size (Figure 1b). Complexes at an N/P ratio 6.7 exhibited in all cases larger average diameters, which suggest that the ability of these polymers to condense ODN at this N/P ratio is impaired. Interestingly, PEG-PEI copolymers with 100 PEG chains of 550 Da exhibited complex sizes, which were comparable to those of the homopolymer PEI 25 kDa. PEG-PEI complexes with ODN are generally found to yield small diameters and narrow size distributions (15). Vinogradov et al. described a copolymer composed of a PEI 2 kDa grafted with an average number of 1.7 PEG (8 kDa) chains, which produced complexes with a phosphorothioate (PS) ODN of 32 nm (7). While experimental conditions and especially chemical modifications of the ODN seem

680 Bioconjugate Chem., Vol. 15, No. 4, 2004

Figure 1. Hydrodynamic diameters of ODN-PEI-25 kDa and block copolymer complexes (a) of the Series 1 and (b) of Series 2 at N/P ratios 6.7 and 20. Values are presented as the mean of 10 runs ( standard deviation.

to have a great influence on complex formation of PEI 25 kDa complexes (31), further studies are necessary to determine the influence of chemical modification on the complexation with block copolymers. Laser Doppler Anemometry (LDA). The surface charge of PEG-PEI complexes was determined at the two different N/P ratios, 6.7 and 20, using LDA. The results are shown in Figure 3. PEI 25 kDa complexes revealed high positive surface charges of +35 mV at N/P 6.7 and +25 mV at N/P 20. A decrease in the zeta potential with increasing N/P ratio has been previously reported for plasmid PEI complexes, as well (26, 32). Zeta potentials of the block copolymer complexes were reduced to nearly neutral values as documented for both series 1 (Figure 2a) and 2 (Figure 2b). Our results are compatible with a core-shell structure, in which hydrophilic PEG chains are oriented toward the surface. Hence, the positive charge of the PEI core is shielded as previously discussed (10, 15). Full shielding was already achieved at the lowest degree of PEGylation investigated in this study of PEG(5k)2 or PEG(550)18, respectively. A further increase of the PEGylation degree had only marginal effects on the zeta potential. The net charge of plasmid DNA complexes could be fully shielded only with block copolymers containing more than six PEG (5 kDa) chains or one very long PEG chain of 20 kDa (26). The substantial difference in DNA size seems to be responsible for the lower degree of condensation of plasmid DNA within the polycationic core (15). Surprisingly, under our experimental conditions only bPEI(25k)-g-lPEG(550)50 led to a clear positive surface charge possibly caused by a difference in stoichiometry of ODN to polymer which is also reflected by the increased complex size.

Brus et al.

Figure 2. Zeta potentials of ODN-PEI 25 kDa and block copolymer complexes (a) of the Series 1 and (b) of Series 2 at N/P ratios 6.7 and 20. Values are presented as the mean of 10 runs ( standard deviation.

Atomic Force Microscopy (AFM). To visualize the shape of the PEG-PEI complexes, we performed atomic force microscopy studies with ODN-PEI 25 kDa complexes and the block copolymer, bPEI(25k)-g-lPEG(550)100 (Figure 3). AFM imaging was performed directly in solution; however, a lower salt concentration of 10 mM NaCl, pH 7.4, was necessary to prevent salt crystal formation (26). Complexes were imaged at an N/P ratio of 6.7. PEI 25 kDa showed images of very small, compact, homogeneously distributed complexes with a nearly spherical shape (Figure 3a,b). PEGylation affected the condensation of the complexes, as is shown for the block copolymer complexes of bPEI(25k)-g-lPEG(550)100 (Figure 3c,d). This has been previously reported for plasmid DNA complexes as well (26). The complexes lost their spherical, well-defined shape and appeared to be more “fuzzy”. Wolfert et al. reported a similar phenomenon for PEGylated PLL (8). The reason for the decreased condensation efficiency of PEGylated derivatives is not yet clear. It might be the result of a reduced ability to collapse into small particles, due to hydrogen bonding of the PEG with water (8, 15). Nuclease Stability. It is well established that complexation with polycations leads to protection of DNA from enzymatic digestion (6, 13). PEG-PEI have demonstrated a high protection efficiency for ODN (7, 9). However, it remains to be investigated whether the degree of PEGylation and PEG molecular weight have an impact on the stabilization efficiency of block copolymers. Both ODN complex series were incubated with

Delivery System for Oligonucleotides and Ribozymes

Bioconjugate Chem., Vol. 15, No. 4, 2004 681

Figure 3. AFM images of ODN complexes of PEI 25 kDa (a and b) in two different resolutions and bPEI(25k)-g-lPEG(550)100 (c and d) in two different resolutions at N/P ratio 6.

DNase 1 for 3 h. Afterward the nuclease was denatured and the ODN displaced from the complex. DNase 1 is a double-strand-specific endonuclease, which also cleaves single stranded ODN, albeit more slowly. Accordingly, nearly 75% of the noncomplexed control was degraded after 3 h in contrast to plasmid DNA which is totally degraded within this period. The percentage of nondegraded control was subtracted from the intact amount of the single samples in order to obtain the real protected fraction, as shown in Figure 4. All block copolymers revealed a protection equal to that of PEI 25 kDa. In contrast, Dash et al. achieved a higher protection of plasmid DNA by a PEGylated PLL in comparison to the non-PEGylated derivative (33). This effect could be explained by steric hindrance of the nucleolytic attack by the PEG shell (7, 9). Under our experimental conditions, however, no specific trend could be distinguished within both series 1 (Figure 4a) and 2 (Figure 4b), suggesting a more complete protection of the ODN by entrapment within the polycationic core. Similarly, Harada et al. reported that an increased degree of polymerization of PEG-PLL block copolymers positively influenced the protection efficiency, which likewise provided evidence for protection via the polycationic segment (9). Complement Activation. The interaction of interpolyelectrolyte complexes with blood and blood components seems to be a major obstacle for their intravenous administration (17, 34). Components of the complement system, as part of the nonadaptive immune system of vertebrate species, have been demonstrated to interact with non viral vectors (24). The binding of complement proteins to DNA complexes leads inevitable to complement-dependent phagocytosis as demonstrated for liposomes (25). Clearance by the reticuloendothelial system reduces the complex half-life within the blood stream. Plank et al. studied the complement activation of commonly used nonviral vectors (24). They reported that a high degree of PEGylation decreased complement activa-

Figure 4. Nuclease stability of ODN-PEI 25 kDa and block copolymer complexes (a) of Series 1 and (b) of Series 2, at N/P ratio 20. All experiments were run in triplicate. (One-way ANOVA statistic test revealed no significant difference).

tion. In this study, the influence of the PEGylation degree, as well as that of PEG molecular weight, on complement activation was systematically investigated. Dilutions of the two series of ODN block copolymer complexes were examined according to the protocol of Plank et al. (24). The appropriate initial serum dilution required for 50% hemolysis was determined and expressed as CH50max. In theory, binding of complement components to the PEG-PEI/ODN complexes during incubation with serum results in a consumption of these complement factors. Thus, they are no longer available for the following hemolysis, leading to a concentration dependent decrease of CH50max. The PEG-PEI complexes of the first series consumed complements component to the same extent as PEI 25 kDa (Figure 5a). Thus, it may be concluded that the shielding obtained by PEGylation with 5 kDa PEG does not necessarily result in decreased complement activation. Increasing PEGylation up to 77% PEG content did not show any positive effects. This is in line with the results of Plank et al. who did not observe a decrease in complement activation of PEGylated PLL when investigating copolymers with up to 20 PEG chains and molecular weights ranging from 1 to 12 kDa. Only larger PEGylation degrees finally led to decreased complement activation (24). On the other hand, PEGylation with 550 Da PEG (Series 2) achieved decreased complement activation (Figure 5b). A correlation between the increase in CH50max and the PEGylation degree was observed. The fact that the PEG-PEIs with 100 chains of PEG 550 Da and 15 chains of PEG 5 kDa contain approximately the same amount of PEG, suggests that not the total PEG content, but rather the PEG molecular weight and, hence, the degree of substitution influences complement activa-

682 Bioconjugate Chem., Vol. 15, No. 4, 2004

Brus et al.

Figure 5. Complement activation of ODN-PEI 25 kDa and block copolymer complexes (a) of Series 1 and (b) of Series 2, at N/P ratio 20. All experiments were run in triplicate.

tion. We hypothesize that a high degree of substitution leads to a brushlike PEG coverage of the polycationic surface. This “crew-cut” micelle structure, achieved only when using the small 550 Da PEG chains, is better capable of preventing complement activation. Movement of the smaller number of longer PEG chains possibly enables contact with the polycationic core. The zeta potential values of all block copolymer complexes suggest that simple shielding of the surface charge is not sufficient to prevent complement activation. Biological Activity of Ribozymes. Ribozymes are short single-stranded RNA-ODN capable of cleaving their target mRNA in a sequence-specific manner (35). Due to the RNA structure, these molecules are even more accessible to enzymatic degradation than the corresponding DNA-ODN (36). It could be previously demonstrated that PEI is able to stabilize ribozymes even under in vivo conditions (37). The ribozyme-based assay used in this study, therefore, was designed to investigate the influence of PEI-g-PEG block copolymer structure on two effects simultaneously: First, the ability to protect ribozymes under cell culture conditions and, second, the biological activity of the ribozyme, therefore, representing very drastic conditions. Ribozymes directed against the luciferase mRNA in stably transfected SKOV-3 cells cause sequence-specific cleavage of the target mRNA, resulting in a decrease of luciferase activity. When cells were incubated with noncomplexed ribozyme as a control, no ribozyme activity could be observed, due to degradation of the ribozyme by the 10% FCS supplement in the cell culture medium. Luciferase activity was comparable to untreated cells, as seen in Figure 6a. Also an unrelated ribozyme, not complementary to the luciferase mRNA, did not affect luciferase activity. In contrast, exposure of cells to ribozyme PEG-PEI complexes led to approximately a 50% down-regulation of the luciferase activity. The interpolyelectrolyte complexes of series 1 (Figure 6b) as well as of series 2 (Figure 6c), demonstrated ribozyme activity comparable to that of PEI 25

Figure 6. Biological activity (a) as ng luciferase/mg protein of ribozyme block copolymer complexes at the N/P ratio of 20 in comparison to unrelated ribozyme (* complexed to bPEI(25k)g-lPEG(5k)6) and noncomplexed ribozyme and as relative light units (RLU) of ribozyme block copolymer complexes (b) of Series 1 and (c) of Series 2, at the N/P ratios 6.7 and 20 in comparison to PEI 25 kDa. All experiments were run in triplicate. Oneway ANOVA statistic test revealed no significant difference between biological activity of PEI 25 kDa complexes and block copolymer complexes.

kDa. Furthermore, no influence of increasing PEGylation on the activity could be distinguished. Surprisingly, the N/P ratio of 6 and 20, respectively, had only a marginal effect on ribozyme activity. In contrast to our observations with ribozyme, it has been reported that the delivery of plasmid DNA using PEG modified PEI often caused a dramatic decrease of transfection efficiency (38, 39). Recently, the transfection activity of five different PEI-g-PEG block copolymers used for plasmid delivery was studied systematically (26). It was found that trans-

Delivery System for Oligonucleotides and Ribozymes

fection was negatively influenced by an increasing degree of PEGylation. This was explained by the differences in cytotoxicity and zeta potential of the various complexes. Under our experimental conditions, only marginal differences in the zeta potentials of the different PEI-g-PEG block copolymer complexes could be observed, which correlates well with the similar biological activity observed. It was speculated that the small ribozyme could be condensed more effectively and the shielding of the positive surface charges could be achieved with lower degrees of PEGylation. It has been reported that the transfection efficiency of PEG-PEI complexes with pSVβ-gal plasmid was slightly decreased with an increase in PEG content. However, PEGylation did not affect the transfection efficiency of the complex for 48 h transfection in the presence of serum (40), analogous to the observation period in this study. Only limited results are available reporting successful ODN transfection using block copolymers (14, 21, 22). An antisense effect directed against amphiphysin was achieved at a very low ODN concentration using a PEGylated polyspermin (22). However, this provides evidence that PEG-PEI copolymers do have potential for an efficient ODN delivery. Under our experimental conditions, ODN transfer was as effective as with PEI 25 kDa. This trend might be further increased using even higher N/P ratios, as demonstrated for plasmid DNA. bPEI(25k)-g-lPEG(550)35 at N/P 50 achieved transfection rates 1 order of magnitude higher than that of PEI 25 kDa at N/P 20, possibly due to the decreased toxicity of PEGylated PEI derivatives (26). Whether this trend in transfection efficiency of plasmid DNA and ODN are comparable remains to be resolved. CONCLUSIONS

The results of this study demonstrate that PEI-g-PEG block copolymers formed very small, homogeneously distributed nucleic acid complexes with high nuclease protection and transfection efficiency. Block copolymers with a high number of short PEG chains (550 Da) achieved a decreased activation of the complement system without losses in biological activity. Therefore, the favorable characteristics of highly grafted copolymers with low molecular weighted PEG are suitable for the delivery of ODN and ribozymes. These new insights into the structure-property relationships of ODN copolymer complexes are valuable for the optimization of hydrophilic nonviral carriers for an efficient ODN delivery. ACKNOWLEDGMENT

The authors thank Beata Urban-Klein for expert assistance with SKOV-3 cell culture. LITERATURE CITED (1) Crooke, S. T. (1998) An overview of progress in antisense therapeutics. Antisense Nucleic Acid Drug Dev. 8, 115-122. (2) Dass, C. R. (2002) Vehicles for oligonucleotide delivery to tumors. J. Pharm. Pharmacol. 54, 3-27. (3) Haseloff, J., and Gerlach, W. L. (1988) Simple RNA enzymes with new and highly specific endo ribonuclease activities. Nature 334, 585-591. (4) Zelphati, O., and Szoka, F. C., Jr. (1996) Intracellular distribution and mechanism of delivery of oligonucleotides mediated by cationic lipids. Pharm. Res. 13, 1367-1372. (5) Merdan, T., Kopecek, J., and Kissel, T. (2002) Prospects for cationic polymers in gene and oligonucleotide therapy against cancer. Adv. Drug Delivery Rev. 54, 715-758. (6) Kabanov, A. V., and Kabanov, V. A. (1995) DNA complexes with polycations for the delivery of genetic material into cells. Bioconjugate. Chem. 6, 7-20.

Bioconjugate Chem., Vol. 15, No. 4, 2004 683 (7) Vinogradov, S. V., Bronich, T. K., and Kabanov, A. V. (1998) Self-assembly of polyamine-poly(ethylene glycol) copolymers with phosphorothioate oligonucleotides. Bioconjugate. Chem. 9, 805-812. (8) Wolfert, M. A., Schacht, E. H., Toncheva, V., Ulbrich, K., Nazarova, O., and Seymour, L. W. (1996) Characterization of vectors for gene therapy formed by self-assembly of DNA with synthetic block copolymers. Hum. Gene Ther. 7, 21232133. (9) Harada, A., Togawa, H., and Kataoka, K. (2001) Physicochemical properties and nuclease resistance of antisenseoligodeoxynucleotides entrapped in the core of polyion complex micelles composed of poly(ethylene glycol)-poly(L-lysine) block copolymers. Eur. J. Pharm. Sci. 13, 35-42. (10) Kataoka, K., Togawa, H., Harada, A., Yasugi, K., Matsumoto, T., and Katayose, S. (1996) Spontaneous Formation of Polyion Complexes Micelles with Narrow Distribution from Antisense Oligonucleotides and Cationic Block Copolymer in Physiological Saline. Macromolecules 29, 8556-8557. (11) Bloomfield, V. A. (1996) DNA condensation. Curr. Opin. Struct. Biol. 6, 334-341. (12) Dunlap, D. D., Maggi, A., Soria, M. R., and Monaco, L. (1997) Nanoscopic structure of DNA condensed for gene delivery. Nucleic Acids Res. 25, 3095-3101. (13) Boussif, O., Lezoualc’h, 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. (14) Kabanov, A. V., Vinogradov, S. V., Suzdaltseva, Y. G., and Alakhov, V. (1995) Water-soluble block polycations as carriers for oligonucleotide delivery. Bioconjugate Chem. 6, 639-643. (15) Seymour, L. W., Kataoka, K., and Kabanov, A. V. (1998) Cationic block copolymers as self-assembling vectors for gene delivery. Self-assembling Complexes for Gene Delivery. From Laboratory to Clinical Trial. ( Kabanov, A. V., Felgner, P. L., Seymour, L. W., Eds.) pp 219-239, John Wiley & Sons, New York. (16) Kabanov, V. A., and Kabanov, A. V. (1998) Interpolyelectrolyte and block ionomer complexes for gene delivery: physicochemical aspects. Adv. Drug Delivery Rev. 30, 49-60. (17) Wiethoff, C. M., and Middaugh, C. R. (2003) Barriers to nonviral gene delivery. J. Pharm. Sci. 92, 203-217. (18) Read, M. L., Dash, P. R., Clark, A., Howard, K. A., Oupicky, D., Toncheva, V., Alpar, H. O., Schacht, E. H., Ulbrich, K., and Seymour, L. W. (2000) Physicochemical and biological characterisation of an antisense oligonucleotide targeted against the bcl-2 mRNA complexed with cationic-hydrophilic copolymers. Eur. J. Pharm. Sci. 10, 169-177. (19) Deshpande, M. C., Garnett, M. C., Vamvakaki, M., Bailey, L., Armes, S. P., and Stolnik, S. (2002) Influence of polymer architecture on the structure of complexes formed by PEGtertiary amine methacrylate copolymers and phosphorothioate oligonucleotide. J. Controlled Release 81, 185-199. (20) Maruyama, A., Katoh, M., Ishihara, T., and Akaike, T. (1997) Comb-type polycations effectively stabilize DNA triplex. Bioconjugate Chem. 8, 3-6. (21) Roy, S., Zhang, K., Roth, T., Vinogradov, S., Kao, R. S., and Kabanov, A. (1999) Reduction of fibronectin expression by intravitreal administration of antisense oligonucleotides. Nat. Biotechnol. 17, 476-479. (22) Mundigl, O., Ochoa, G. C., David, C., Slepnev, V. I., Kabanov, A., and De Camilli, P. (1998) Amphiphysin I antisense oligonucleotides inhibit neurite outgrowth in cultured hippocampal neurons. J. Neurosci. 18, 93-103. (23) Vinogradov, S., Batrakova, E., Li, S., and Kabanov, A. (1999) Polyion complex micelles with protein-modified corona for receptor-mediated delivery of oligonucleotides into cells. Bioconjugate. Chem. 10, 851-860. (24) 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, 1437-1446. (25) Wassef, N. M., and Alving, C. R. (1993) Complementdependent phagocytosis of liposomes. Chem. Phys. Lipids 64, 239-248.

684 Bioconjugate Chem., Vol. 15, No. 4, 2004 (26) 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, 845-854. (27) Petersen, H., Fechner, P., Fischer, D., and Kissel, T. (2002) Synthesis, Characterization, and Biocompatibility of Polyethylenimine-graft-poly(ethylene glycol) Block Copolymers. Macromolecules 35, 6867-6874. (28) Kissel, T., Petersen, H., Fischer, D., Kunath, K., and von Harpe, A. (1999) Kationische Blockcopolymere. DE19933024A1 (EP00/06214/WO01/05875A1), Germany. (29) Whaley, K., and North, J. (1997) Hemolytic assays for whole complement activity and individual components. Complement. A practical approach (Dodds, A. W., and Sim, R. B., Eds.) pp 19-47, Oxford University Press, New York. (30) Hill, H. D., and Straka, J. G. (1988) Protein determination using bicinchoninic acid in the presence of sulfhydryl reagents. Anal. Biochem. 170, 203-208. (31) Dheur, S., Dias, N., van Aerschot, A., Herdewijn, P., Bettinger, T., Remy, J. S., Helene, C., and Saison-Behmoaras, E. T. (1999) Polyethylenimine but not cationic lipid improves antisense activity of 3′-capped phosphodiester oligonucleotides. Antisense Nucleic Acid Drug Dev. 9, 515-525. (32) Kunath, K., Harpe von, A., Petersen, H., Fischer, D., Voigt, K., Kissel, T., and Bickel, U. (2002) The Structure of PEGModified Poly(ethylen imines) Influences Biodistribution and Pharmacokinetics of their Complexes with NF-kB Decoy in Mice. Pharm. Res. 19, 810-817. (33) Dash, P. R., Toncheva, V., Schacht, E., and Seymour, L. W. (1997) Synthetic polymers for vectorial delivery of DNA: characterisation of polymer-DNA complexes by photon correlation spectroscopy and stability to nuclease degradation

Brus et al. and disruption by polyanions in vitro. J. Controlled Release 48, 269-276. (34) Kircheis, R., Wightman, L., and Wagner, E. (2001) Design and gene delivery activity of modified polyethylenimines. Adv. Drug Delivery Rev. 53, 341-358. (35) Birikh, K. R., Heaton, P. A., and Eckstein, F. (1997) The structure, function and application of the hammerhead ribozyme. Eur. J. Biochem. 245, 1-16. (36) Scherr, M., Grez, M., Ganser, A., and Engels, J. W. (1997) Specific hammerhead ribozyme-mediated cleavage of mutant N-ras mRNA in vitro and ex vivo. Oligoribonucleotides as therapeutic agents. J. Biol. Chem. 272, 14304-14313. (37) Aigner, A., Fischer, D., Merdan, T., Brus, C., Kissel, T., and Czubayko, F. (2002) Delivery of unmodified bioactive ribozymes by an RNA-stabilizing polyethylenimine (LMWPEI) efficiently down-regulates gene expression. Gene Ther. 9, 1700-1707. (38) 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. (39) Erbacher, P., Bettinger, T., Belguise-Valladier, P., Zou, S., 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, 210-222. (40) Choi, J. H., Choi, J. S., Suh, H., and Park, J. S. (2001) Effect of Poly(ethylene glycol) Grafting on Polyethylenimine as a Gene Transfer Vector in vitro. Bull. Korean Chem. Soc. 22, 46-52.

BC034160M