548
Bioconjugate Chem. 2002, 13, 548−553
Aliphatic Ionenes as Gene Delivery Agents: Elucidation of Structure-Function Relationship through Modification of Charge Density and Polymer Length Alexander N. Zelikin,†,‡ David Putnam,§ Prasad Shastri,§ Robert Langer,*,§ and Vladimir A. Izumrudov*,† M. V. Lomonosov Moscow State University, Chemistry Department, Vorob'evi Gori V-234, 119899, Moscow, Russia; Massachusetts Institute of Technology, Division of Health Sciences and Technology, E25-342, Cambridge, Massachusetts, 02139; and Massachusetts Institute of Technology, Chemical Engineering Department, E25-342, Cambridge, Massachusetts, 02139. Received September 6, 2001; Revised Manuscript Received November 29, 2001
Polycation-based gene delivery agents are generally polydisperse populations whose properties are averaged among the different molecular weight species. Therefore, to understand the physicochemical properties of polycations and their relationships to cellular gene transfer, one needs to control the molecular weight of the polymer as well as its cationic charge density. To investigate the structurefunction correlation of polycations with respect to the degree of polymerization (DP) and charge density, a series of model materials based on aliphatic ionenes was synthesized and fractionated into distinct molecular weight fractions with DP range from 14 to 32. The aliphatic ionene fractions and their polyelectrolyte complexes (PEC) with DNA were studied using physicochemical and biological methods. Ionene polymers were shown to possess low cytotoxicity (minimal viability of the P388D1 murine macrophage cells 80%). DP and charge density of the ionenes were shown to be the factors of effective control of PEC dissociation in water-salt solutions, with a diminished role of charge density upon lengthening the ionene chain. These polymer characteristics were also important for DNA-ionene PEC resistivity to DNase activity and the ability of ionenes to serve as gene delivery vectors in vitro and exhibited good correlation with the results of salt-induced dissociation of PEC. These data may be useful for developing correlations and mathematical models to predict synthetic gene delivery vector efficiency.
INTRODUCTION
With the advent of modern molecular biology and the potential to not only identify the genetic basis of disease, but also potentially treat diseases at the genetic level, came the promising field of gene therapy. However, while new targets for gene therapy are discovered at an extraordinary pace, the discovery of safe and effective gene delivery vehicles has lagged behind. Currently, there are approximately 400 gene therapy protocols in clinical trials worldwide (see http://www.wiley.co.uk/ genetheraphy/clinical/coutries.html) and the success of these clinical trials is directly influenced by the efficiency of the delivery system. The most effective vectors for transfection are those based on viruses and viral peptide sequences (1). However utilization of these agents is substantially hindered by immunologic issues. Therefore attempts are being made to create nonviral gene-delivery agents, such as liposomes and polycations (1). The latter effectively bind to negatively charged DNA, which results in substantial * To whom all correspondence should be addressed: phone (+ 7 095) 9393117, fax (+7 095) 9390174, email: izumrud@ genebee.msu.su and phone (671) 2533413, fax (617) 2588827, email:
[email protected]. † M. V. Lomonosov Moscow State University. ‡ Massachusetts Institute of Technology, Division of Health Sciences and Technology. § Massachusetts Institute of Technology, Chemical Engineering Department.
DNA compaction. Moreover, if the polyelectrolyte complex (PEC) has an overall positive charge it significantly increases the interaction of PEC with negatively charged cell membranes. Different groups have studied DNApolycation interactions with respect to the delivery of DNA in vitro and in vivo. The structure of the polycation and its affinity to DNA has a substantial influence on transfection level (1-3). While the field of gene transfer with polycationic agents is maturing, to date, general data reporting the correlation between macromolecular characteristics, such as degree of polymerization (DP) and charge density, on the transfection capacity of the polycation are unavailable in the literature. In fact, there is discrepancy in the literature regarding polymer structure and transfection capacity. Godbey et al. (4) studied transfection with complexes of DNA with polyethylenimine (PEI) and found that the level of reporter gene expression increases with increasing PEI molecular weight. Similar results were obtained by Wolfert et al. (5) for systems based on methacrylate-derived polycations. This was also true in the case of polymers with quaternary ammonium ions as positive moieties in the side chain, although the observed transfection levels were rather low (5). In contrast, Schaffer et al. (6) showed that the dependence of transfection efficiency as a function of polylysine DP has a maximum at DP ≈ 36. Other properties of polycations were also shown to be dependent on DP and charge density and sometimes contrast with the efficiency of transfection. For example cytotoxicity of polycations
10.1021/bc015553t CCC: $22.00 © 2002 American Chemical Society Published on Web 03/22/2002
Aliphatic Ionenes as Gene Delivery Agents Scheme 1
increased with growing polymer charge density (7). Plank et al. (8) showed that activation of the complement system by synthetic DNA complexes is a potential barrier for gene delivery and that long chain polylysines are much stronger activators compared to short chain samples. One approach to the design of better polycation-based vectors is to describe the steps of complexation in physicochemical terms. In other words there is a need for mathematical descriptions of the kinetics and thermodynamics of the vector pathway from administration to gene expression, and efforts have been made to create such models (9, 10). However, to date, no model can fully predict the efficiency of a novel vector. Thus detailed studies of the connection between the physicochemical and biological properties of the DNA-containing PEC could be useful. The goal of this study was to investigate the influence of DP and charge density of polymeric quaternary ammonium salts upon complexing with DNA and biological properties of PEC formed in order to correlate physicochemical features of the polycations and its biological activity. To accomplish this we used model polycations with charged quaternized nitrogen atoms in the polymer backbone (Scheme 1). Such polymers constitute the family of polyelectrolytes called ionenes (11). These polymers are attractive in the scope of this study, as the charge density of ionenes is solely determined by the choice of starting materials for synthesis and thus can be changed at will in a wide range. Moreover, with these materials it is possible to vary one parameter while keeping the other constant. For example, it is possible to obtain samples with the same degree of polymerization but different charge density, or to keep the intercharge spacing constant and alter the polymer chain length. EXPERIMENTAL SECTION
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), ethidium bromide, TAE, and HEPES buffer were purchased from Sigma (St. Louis, MO). DNAse I was purchased from USB Corporation. P388D1 mouse macrophage cell line was a kind gift from Zycos (Cambridge, MA). Cell culture was grown in RPMI medium 1640 with 10 vol % of FBS, and the medium also contained penicillin (100 units/mL) and streptomycin (100 µg/mL). Immortalized African green monkey kidney fibroblasts (COS7) were purchased from American Type Culture Collection (Manassas, VA) and grown in Dubelco’s Eagle Modified Medium (DMEM) containing 10 vol % FBS, 100 units/mL penicillin, 100 µg/mL streptomycin. All cell culture growth and incubation was conducted at 37 °C in 5% CO2 atmosphere. Sodium salt of highly polymerized calf thymus DNA (∼10000 base pairs) was purchased from Sigma and used without purification. Concentration of DNA phosphate groups was determined by UV absorbance measurements
Bioconjugate Chem., Vol. 13, No. 3, 2002 549
at 260 nm assuming molar extinction coefficient 260 ) 6500 L mol-1 cm-1. pCMV-Luciferase was purchased from Elim Biopharmaceuticals, Inc. (San Francisco, CA). Polymer Synthesis. 2,4-, 2,8-, and 2,10 ionenes were synthesized, fractionated, and purified as described elsewhere (12). Gel Retardation. DNA-ionene complexes were formed by mixing 50 µL of pCMV-Luciferase DNA solution (0.2 mg/mL in 10 mM HEPES buffer, pH 7.2) and 50 µL of solution of ionene at a desired concentration with vortexing. Complexes were allowed to form for 30 min at room temperature. After that, 20 µL of the solution was loaded onto 1% agarose gel in TAE buffer and ran at 90 V for 60 min. The gels were visualized with ethidium bromide. Complex Size and Zeta-Potential Measurement. Polyelectrolyte complexes were formed as described above for the gel retardation assay. The mixture was diluted to 1.6 mL with HEPES buffer. Complex particle size and zeta potentials were measured at room temperature using Zeta-pals dynamic light scattering detector (Brookhaven Instruments, Holtsville, NY) with a 15 mV laser and incident laser beam at 676 nm. Correlation functions were collected at a scattering angle of 90°. Particle sizes were calculated using multiple angle sizing option of the instruments particle-sizing software. Particle sizes are expressed as effective diameters assuming log-normal distribution. Zeta potentials were calculated using the Smolukowsky model for aqueous suspensions. DNAse Resistivity. Stoichiometric DNA-polycation complexes, prepared as described above, were placed into 1 mL of 10 mM PBS buffer containing 5 mM MgCl2 at 25 °C. To this solution was added 10 units of DNAse I, and optical density of the solution at 260 nm was recorded for 30 min. Fluorescence Quenching. Fluorescence intensity was measured using Jobin-Yvon-3CS Spectrofluorimeter (France) in a water-thermostatic stirred cell holder. The measurements were performed in a quartz fluorescence cell with continuous stirring. The excitation and emission wavelengths were 535 and 595 nm, respectively. Concentration of calf thymus DNA, expressed in concentration of phosphate groups, was [P] ) 4 × 10-5 base-mol L-1 in all experiments. Solutions of polyelectrolyte complexes were prepared directly in the fluorimetric cell by mixing of definite volumes of stock solutions of the polymers. Fluorimetric titration was performed by successive addition of 4 M NaCl solution and measuring the intensity of fluorescence. Time interval between the additions was 5 min. Cytotoxicity. Cytotoxicity of the polymers was evaluated using the MTT assay (13). P388D1 cells were grown in 96-well plates at initial cell density of 10000 cells per well in 0.2 mL of growth medium for 24 h, after which time the media was replaced with growth medium containing polymers at various concentrations and incubated at 37 °C for 24 h. At this time 25 µL of MTT solution (5 mg/mL) in sterile PBS buffer was added to each well and incubated for 2 h, after which 100 µL of extraction buffer (20% w/v SDS in 50:50 v/v DMF-water, pH 4.7) was added to each well and incubated for 24 h at 37 °C. The optical densities were measured using a microplate reader (Dynatech MR5000) at 560 nm. The results are expressed as percent relative to the control (no polymer added) cells. Transfection. Immortalized African monkey kidney fibroblasts (COS-7) were grown in six-well plates at initial densities of 200000 cell per well in 2 mL of growth media for 24 h. After this the growth medium was
550 Bioconjugate Chem., Vol. 13, No. 3, 2002
replaced with 2 mL of reduced serum media and polymer complexes were added to the corresponding wells. Polymer complexes were prepared as described for the gel assay analysis. Each sample contained 10 µg of DNA. After 4 h of incubation the media was changed to DMEM and cells were allowed to grow for 36 h. At this time growth media was removed, and cell lysates were formed and analyzed for luciferase activity (Promega luciferase assay kit) and total protein content (BCA assay, Pierce). RESULTS AND DISCUSSION
The first report of ionenes dates back to the 1930s (14). In the 1940s Rembaum et al. (11) conducted an extensive research on ionene formation kinetics and thermodynamics. In 1961 Kimura et al. (15) isolated three samples of hexadimethrine bromide, 3,6-ionene, with different molecular weights and studied toxicity of these samples on mice. They found that LD50 significantly decreases with the lengthening of the ionene chain length. Efimov et al. (16) performed similar experiments and showed that the LD50 is also dependent on the charge density of ionenes: the longer the intercharge spacing along the ionene chain the lower the value of LD50. To the best of our knowledge, the only report of the use of ionene polymers for gene delivery is that of Aubin et al. who used 3,6-ionene for DMSO-assisted transfection (17). However this study was focused on optimization of DMSO-assisted transfection, rather than influence of the ionene characteristics on its ability to induce gene delivery. Yet until recently there was no efficient way to fractionate polydisperse ionene samples. Thus the experimental results could be masked by broad molecular weight distribution of ionene samples. Recently (12) we described a method of effective fractionation of polydisperse samples of ionenes, which allowed us to obtain samples of 2,4-, and 2,8-ionenes in the range of DP 10 to 30 and study the physicochemical properties of ionene-containing polyelectrolyte complexes with poly(methacrylic acid). In this paper we investigate the physicochemical properties of DNA-containing PEC with ionenes and correlate them to ability of the ionenes to serve as gene delivery agents. For simplicity a sample of x,y-ionene with a certain DP will be denoted as ionene[x,y/DP], where DP equals to the number of charged nitrogen atoms per ionene chain. In all cases the charge ratio of the polyelectrolytes is expressed in the ratio of the base-mole concentrations of charged moieties, φ t [+]/[-]. Complexes of plasmid DNA (pCMV-Luciferase) with ionenes were formed at different charge ratios of polyelectrolytes, and agarose gel electrophoresis was subsequently performed. This method allows visualization of the interaction of DNA and polycations. Uncomplexed or incompletely neutralized DNA, φ < 1, migrates in the electric field toward the anode, whereas stoichiometric complexes or complexes with an excess of positive charges do not migrate. Staining the gel with ethidium bromide (EB) provides a picture of the complexes’ position on the gel. A representative picture for different ionenes is presented in Figure 1. As anticipated, in all the studied cases the complete retardation of DNA migration is observed at φ g 1. The same complexes were analyzed for size using dynamic light scattering. The complex particle size as a function of ionene-DNA charge ratio for ionene-[2,4/20] is plotted in Figure 2. The increase of ionene content is accompanied by a substantial decrease in PEC particle size. This is true up to the charge ratio of unity, where the particles are neutral and form aggregates. With
Zelikin et al.
Figure 1. Electrophoretic retardation of pCMV-luciferase by ionene-[2,4/20]. Numbers signify the ionene-DNA charge ratios.
Figure 2. Dependence of the DNA-ionene-[2,4/20] complex particle size on the charge ratio of the polyelectrolytes.
excess polycation, charged complex particles do not aggregate, and at charge ratios of the polymers 2:1 the particle size is ∼100 nm. This means that DNA-ionene[2,4/20] complexes meet the requirement to form complexes less than 150 nm diameter, making endocytosis feasible (18). Analogous dependencies were obtained for all utilized ionene samples (data not shown). All ionenes formed complexes with DNA with a particle size about 100 nm at charge ratios of the polymers greater than 2:1 regardless of charge density or chain length of the polycation. This correlates well with the results of Wolfert et al. (5), who showed, that regardless of DP and the nature of the charged moiety, polycations can complex DNA into small particles. With excess polycation, soluble complex particles are thought to have an overall positive charge. To investigate this with ionenes, zeta potentials of the complex particles were determined (Figure 3). At φ > 1 the surface charge of DNA-ionene complex particles is positive, as is the case for all studied samples of ionenes regardless of polycation chain length and charge density. The incorporation of DNA into a PEC can substantially hinder its accessibility to DNAse and thus dramatically decrease nuclease activity. Figure 4 shows the results of experiments in which DNA (curve 1) and DNA-ionene [2,8/29] complex φ ) 1 (curve 2) were treated with DNAse I. Nuclease activity was monitored by the change in optical density at 260 nm (19). These data show that accessibility of nucleic acid to DNAse is much lower in the case of PEC compared to uncomplexed DNA. Dependencies analogous to those presented in Figure 4 were obtained for each DNA-ionene complex and the ∆A260 (∆t ) 30 min) was calculated (Figure 5). On this plot the greater the value of 1/∆A260, the more stable the
Aliphatic Ionenes as Gene Delivery Agents
Bioconjugate Chem., Vol. 13, No. 3, 2002 551
Figure 3. Zeta potentials of DNA-containing complexes with ionene-[2,4/20] (1), [2,8/14] (2), [2,8/20] (3), [2,8/29] (4) and [2,10/20] (5) at different charge ratios of the polyelectrolytes, φ ) [+]/[-].
Figure 5. DNAse resistivity for stoichiometric DNA-containing complexes as a function of ionene DP for 2,4-ionene (1), 2,8ionene (2), and 2,10-ionene (3).
Figure 4. Dependence of optical density at 260 nm in solutions of DNAse I and DNA (1) and DNA-ionene-[2,4/29] PEC, φ ) 1 (2); PBS buffer, 0.005 M MgCl2.
Figure 6. Dependence of fluorescence intensity I of the DNAionene-[2,4/20] solution on charge ration of the polymers, φ ) [+]/[-]. [P] ) 4 × 10-5 M; [P]/[EB] ) 20/1; Tris buffer, pH 9.0; 25 °C.
complex upon DNAse treatment. Comparing data for different ionenes with the same DP shows that a decrease of ionene charge density is followed by a decrease in the resistance of its complex with DNA to nuclease activity. The increase of ionene chain length increases the ionene’s ability to shield DNA from nuclease. The stability of polyelectrolyte complexes formed by calf thymus DNA and ionenes in water-salt solutions was studied using a fluorescence quenching technique (5, 20) employing ethidium bromide as a fluorescent probe. The results of titration of DNA with ionene-[2,4/ 20] are presented in Figure 6. As seen, the introduction of ionene into the solution of DNA-EB complex leads to a pronounced decrease of fluorescence intensity, I, which suggests the formation of a PEC (20). This is true up to the charge ratio of the polymers equal to unity, where quenching ceases. The dissociation of DNA-ionene complex (φ ) 1.2) in water-salt solutions was determined by NaCl titration in the presence of EB. The fluorescence intensity of the polyelectrolyte mixture I was normalized by I0: intensity
of fluorescence of DNA-EB complex at the same salt concentration. In all experiments the ratio of DNA (expressed in concentration of phosphate groups, [P]) to EB was [P]/[EB] ) 20/1. As a quantitative measure of the stability of PEC to dissociation in water-salt solution, we define the value C*NaCl, which is the value of CNaCl that corresponds to the midpoint of the curve of fluorescence ignition. Figure 7 presents the dependence of C*NaCl vs DP of the ionene for complexes of DNA with 2,4-ionene (curve 1) and 2,8-ionene (curve 2). As follows from curve 1, the stability of DNA-containing complexes with 2,4-ionene significantly increases with the lengthening of the chain of polycation. Likewise, C*NaCl for the complex DNA-ionene-[2,4/8] is 0.13 mol L-1, whereas for the complex with ionene-[2,4/20] this value is equal to 0.27 mol L-1. It is noteworthy that in the region of the shortest chains of 2,4-ionene the stability of DNA-containing complexes decreases particularly rapidly. This correlates well with the concept that the value of cooperative binding length for most
552 Bioconjugate Chem., Vol. 13, No. 3, 2002
Figure 7. Dependence of C*NaCl, M on the degree of polymerization of the ionene in solutions of PEC formed by DNA with 2,4-ionene (1) and 2,8-ionene (2). φ ) 1,2; other conditions as in Figure 6.
examined pairs of polyelectrolytes lies in the range 4 to 7 monomer units (21). The dependence obtained for the complexes of DNA with 2,8-ionenes of different DP (Figure 7, curve 2) is similar to that for DNA-2,4-ionenes. In this case the growth of the stability of the complex with the increase of polycation chain length is even more pronounced. For short chains of ionenes the stability of 2,4-ionene containing complexes is higher than for the complexes with 2,8-ionene, as would be expected from higher charge density of 2,4-ionene (12, 20). Strikingly, upon lengthening of the ionenes chains the difference in stability of the complexes decreases, and on attainment of 20-25 monomer units, the difference between the ionenes as partners for complexation with DNA becomes vanishingly small. In other words, the charge density proved to be the factor controlling stability of DNA-containing PEC only in the case of the ionenes with relatively low DPs. The cytotoxicity of the ionenes was studied using P388D1 macrophages (Figure 8). The minimum viability of the cells exposed to solutions of ionenes at various concentrations of polycation is 80% relative to controls, indicating low cytotoxicity of ionenes compared to polyethyleneimine and polylysine (22), which is critical to normalize the polymer effects in the cell-based transfection assays. As the whole range of viability of cells treated with different ionenes is within experimental error, it is difficult to deduce the influence of the DP and intercharge spacing of ionenes on their toxicity. However, our data indicate that the influence of these parameters on sample cytotoxicity in vitro is substantially lower than that found in vivo (15, 16). The data obtained also suggest that the cytotoxicity is a function of the nature of the polycation charged moiety (primary, secondary, tertiary, or quaternary amine group) rather than the DP of the macromolecule or its charge density. Polymeric quaternary ammonium salts possess little activity as gene delivery agents. While most of the polycations with different types of charged moieties are able to condense DNA into small particles with an overall positive surface charge, the efficiency of transfection dramatically drops from polylysine-based vectors to quaternary ammonium polymers (5).
Zelikin et al.
Figure 8. In vitro cytotoxicity as a function of ionene basemole concentration for ionenes [2,4/20] (1), [2,8/14] (2), [2,8/20] (3), [2,8/29] (4) and [2,10/20] (5).
Figure 9. pCMV-Luciferase expression as a function of ionene DP for 2,4-ionene (1), 2,8-ionene (2), and 2,10-ionene (3).
Nevertheless, even this relatively low activity allows examining the ability of ionenes to induce nucleic acid delivery with respect to the influence of polycation macromolecular characteristics. It is well documented that most polymer properties change rapidly in the region of low DP values, while at a certain DP one would expect a property-structure plateau (23). For this reason we studied low DP trends using aliphatic ionenes (Figure 9). To determine the optimal charge ratio of the polymers for transfection, COS-7 cells were exposed to DNAionene-[2.4/20] complexes with different ionene contents. These experiments showed that the highest levels of reporter gene expression are observed at a charge ratio 4:1 (data not shown). In all further experiments, cells were treated with DNA-ionene complexes at φ ) 4. Using samples of ionenes with the same charge density and different chain lengths revealed the influence of polycation chain length on the transfection level. Figure
Aliphatic Ionenes as Gene Delivery Agents
9 shows that increase of chain length of the ionene leads to a pronounced increase of reporter gene expression. This observation is true for both 2,4- and 2,8-ionene. This result corresponds well with those obtained for synthetic gene delivery vectors based on polyethyleneimine (4) and other polycations (5). It is also consistent with the data obtained by Schaffer et al. (6), who showed that the transfection level as a function of polylysine DP has a maximum at DP ≈ 36. As stability of DNA-ionene complexes is lower than that with polylysine (20), one can expect the appearance of such maximum in case of ionene-based vectors at DPs higher than 36 monomer units. Unfortunately, we could not isolate ionene samples with DP higher than 30; this is a subject of ongoing research. To scrutinize the influence of polymer charge density we performed transfections with complexes of DNA with ionenes of the same chain length but different intercharge distance, ionenes [2.4/20], [2.8/20], and [2.10/20]. Data, presented in Figure 9, show that the increase of ionene charge density is followed by a pronounced increase in transfection level. Likewise, in the case of 2,4ionene, the observed value of RLU/mg is ∼1,5 times higher than for 2,8-ionene. For the latter, in turn, it is 2 times higher than for 2,10-ionene, for which the level of transfection is only slightly higher than that observed for uncomplexed DNA. Another interesting notion comes from comparing the data in Figures 5, 7, and 9. The influence of ionene charge density and chain length on PEC stability in water-salt solutions and tolerance to DNAse action correlate well with transfection efficiency. Moreover, it can be observed that different ionenes which complexes with DNA possess similar stability to dissociation in water-salt solutions induce equivalent DNA expression. This finding shows the strong influence of polymer microstructure and molecular weight on the PEC properties. According to the revealed correlation, physicochemical results may be useful for predictions of PEC applicability for transfection. The revealed extraordinary correlation between the results obtained by different approaches (stability of the complexes toward dissociation in water-salt solutions, resistivity toward DNase restriction and transfection efficiency) certainly does not imply that the results of different approaches are directly transferable, yet it shows that physicochemical evaluation of the system can give a reasonable prediction of the biological usefulness of the system and can be used for creating mathematical models for prediction of a novel vector. ACKNOWLEDGMENT
Authors would like to thank Dr. D. Lynn, Dr. D. Anderson, Dr. C. Gentry, and D. Ting for significant
Bioconjugate Chem., Vol. 13, No. 3, 2002 553
contribution to the work. Financial support of the study by NIH grant # 2-R01-6M26698-21, BPEC grant # EEC9843342 and Russian Fund of Fundamental Research 9903-33399a are gratefully acknowledged. LITERATURE CITED (1) Kabanov, A. V., Felgner, P. L., and Seymour, L. W. (1998) Self-Assembling Complexes for Gene Delivery: From Laboratory to Clinical Trial, John Wiley and Sons, Inc., New York,. (2) Varga, C. M., Wickham, T. J., and Lauffenburger, D. A. (2000) Biotechnol. Bioeng. 70, 593-605. (3) Luo, D., and Saltzman, W. M. (2000) Nature Biotechnol. 18, 33-37. (4) Godbey, W. T., Wu, K. K., and Mikos, A. G. J. Biomed. Mater. Res. (1999) 45, 268-275. (5) Wolfert, M. A., Dash, P. R., Nazarova, O., Oupicky, D., Seymour, L. W., Smart, S., Strhalm, J., and Ulbrich, K. (1999) Bioconjugate Chem. 10, 993-1004. (6) Schaffer, D. V., Fidelman, N. A., Dan, N., and Lauffenburger, D. A. (2000) Biotechnol. Bioeng. 67, 598-606. (7) Plank, C., Zatloukal, K., Cotton, M., Mechtler, K., and Wagner, E. (1992) Bioconjugate Chem. 3, 533-539. (8) Plank, C., Mechtler, K., Szoka, F. C., and Wagner, E. (1996) Human Gene Therapy 7, 1437-1446. (9) Ledley, T. S., and Ledley, F. D. (1994) Human Gene Therapy 5, 679-691. (10) Dee, K. U., and Shuler, M. L. (1997) Biotechnol. Bioeng. 54, 468-490. (11) Rembaum, A., Baumgartner, W., and Eisenberg, A. (1968) Polym. Lett. 6, 159-171. (12) Zelikin, A. N., Akritskaya, N. I., and Izumrudov, V. A. (2001) Macromol. Chem. Phys. 202 (15), 3018-3026. (13) Hansen, M. B., Neilsen, S. E., and Berg, K. (1989) J. Immunol. Methods 22, 203-210. (14) Littmann, E. R., and Marvel, C. S. (1930) J. Am. Chem. Soc. 52, 287-294. (15) Kimura, E. T., Young, P. R., and Barlow, G. H. (1962) Proc. Soc. Exp. Biol. Med. 111, 37-41. (16) Efimov, V. S., Gulaeva, J. G., Menshova, G. I., Razvodovskiy, E. V., Zezin, A. B., and Lakin, K. M. (1974) Pharmakologiya 6, 688. (17) Aubin, R. A., Weinfeld, M., Mirzayans, R., and Paterson, M. C. (1994) Mol. Biotechnol. 1, 29-48. (18) McGraw, T. E., and Maxfield, F. R. In Targeted Drug Delivery, Juliano, R. L. Ed.; Springer: New York, 1991; p 1141. (19) Katayose, S., and Kataoka, K. (1997) Bioconjugate Chem. 8, 702-707. (20) Izumrudov, V. A., Zhiryakova, M. V., and Kudaibergenov, S. E. (2000) Biopolymers 52, 94-108. (21) Papisov, I. M., and Litmanovich, A. A. (1988) Adv. Polym. Sci. 90, 139-179. (22) Putnam, D., Gentry, C. A., Pack, D. W., and Langer, R. (2001) Proc. Nat. Acad. Sci. U.S.A. 98, 1200-1205. (23) Odian, G. Principles of Polymerization, 3rd ed., John Wiley and Sons, Inc., New York, 1991.
BC015553T
