ARTICLE
doi:10.1016/j.ymthe.2004.11.015
Structure/Property Studies of Polymeric Gene Delivery Using a Library of Poly(B-amino esters) Daniel G. Anderson, Akin Akinc, Naushad Hossain, and Robert Langer* Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA *To whom correspondence and reprint requests should be addressed. E-mail:
[email protected].
Available online 24 December 2004
Here we describe the synthesis and characterization of a library of 486 second-generation poly(Bamino esters). To understand better the structure/property relationships governing polymeric gene delivery, we synthesized polymers with 70 different primary structures, at 6 to 12 different molecular weights, using monomers previously identified as common to effective gene delivery polymers. This library was characterized by (1) molecular weight, (2) particle size upon complexation with DNA, (3) surface charge upon complexation with DNA, (4) optimal polymer/ DNA ratio, and (5) transfection efficiency. In this library, polymers with 20 of the 70 primary structures possess transfection efficiencies as good as or better than one of the best commercially available lipid reagents, Lipofectamine 2000. In general, the most effective polymers condense DNA into sub-150-nm complexes with positive surface charge. Among this group, the 2 most effective polymers condensed DNA to the smallest particle sizes (71 and 79 nm). Interestingly, the top 9 polymers were all formed from amino alcohols, and the structure of the 3 top performing polymers differs by only one carbon. This convergence in structure of the top performing polymers suggests a common mode of action and provides a framework with which future polymers can be designed.
INTRODUCTION The safe and effective delivery of DNA remains a central challenge to the application of gene delivery in the clinic. Currently, the majority of gene therapy protocols employ viral delivery systems, which are associated with serious toxicity and production concerns [1]. Nonviral systems offer a number of potential advantages, including ease of production, stability, low immunogenicity and toxicity, and capacity to delivery larger DNA payloads [2]. However, existing nonviral delivery systems are far less efficient than viral vectors [3]. A promising group of nonviral delivery compounds is the cationic polymers, which spontaneously bind and condense DNA. A wide variety of cationic polymers that transfect have been characterized. Some are natural polymers, such as protein [4] and peptide systems [5], while others are synthetic polymers such as poly(ethylene imine) (PEI) [6] and dendrimers [7]. Recent advances in polymeric gene delivery have in part focused on the incorporation of biodegradability to decrease toxicity. Typically, these polymers contain both chargeable amino groups, to allow for ionic interaction with the negatively charged DNA phosphate, and a degradable region, such as a hydrolyzable ester linkage. Several examples of these include poly[a-(4-aminobutyl)-l-glycolic acid] [8], net-
426
work poly(amino ester) [9], and poly(h-amino esters) [10– 14]. Poly(h-amino esters) show particular promise as delivery agents, as they are highly efficient in vitro and easily synthesized via the conjugate addition of a primary amine or bis(secondary amine) to a diacrylate. Until recently, the development of new polymers was an iterative process: polymers were designed one at a time and then individually tested for their properties. Combinatorial approaches have been developed that facilitate the generation of structurally diverse libraries of polymeric biomaterials [15,16]. This combinatorial approach has also been applied to the discovery of gene delivery polymers. Murphy et al. generated a targeted combinatorial library of 67 peptoids via solid-phase synthesis and screened them to identify new gene delivery agents [17]. We recently described the semiautomated, solution-phase parallel synthesis and evaluation of a library of 2350 structurally diverse, degradable poly(h-amino esters) [12]. The high-throughput screening of these materials resulted in the discovery of 47 polymers that transfect as well as or better than the current gold standard, PEI. These studies demonstrate the potential of high-throughput combinatorial approaches to accelerate the discovery of new gene delivery polymers and also identified several structures common in effective polymers.
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy 1525-0016/$30.00
ARTICLE
doi:10.1016/j.ymthe.2004.11.015
In addition to polymer composition, the polymer molecular weight and polymer/DNA ratio are critical for transfection efficiency [10,18–20]. The molecular weight of a polymer is proportional to the number of polymer cation–DNA anion interactions and therefore affects the affinity of the polymer to DNA [20]. Short polymers do not effectively condense DNA, while polymers that are too long retard the bunpackingQ of DNA required for transcription and translation [20]. The polymer/DNA ratio reflects the charge ratio between the positively charged polymer and the negatively charged DNA. This impacts a number of important transfection properties, such as the stability, cellular uptake level, and cytotoxicity of the resulting complex. Recently, we demonstrated that the end-terminal chemistry of poly(h-amino esters) also significantly affects their gene delivery efficacies [10]. Depending on the ratio of monomers during synthesis, poly(h-amino esters) can be made to have either amine-or acrylateterminated chains. In our initial study using two model poly(h-amino ester) polymers, we demonstrated that both transfection and toxicity are significantly affected by the polymer end groups [10]. Using structure/function information gained in our initial work, here we synthesize, in gram scale, and characterize a library of 486 poly(h-amino esters). To understand more completely the structure/function properties governing polymeric gene delivery, these polymers were (1) synthesized at a range of molecular weights, (2) tested in quadruplicate at six different polymer/DNA ratios, (3) complexed to DNA and (4) characterized for particle size and surface charge. The best performing polymer, C32, delivers DNA in vitro
better than the current state-of-the art nonviral delivery systems.
RESULTS Polymer Synthesis Previously, we performed a high-throughput synthesis and screening of a library of poly(h-amino esters) [12]. From this, we identified several monomers as common to effective gene delivery polymers. Amino monomers in effective polymers contained an alcohol, imidazole, or a secondary diamine, while acrylate monomers were almost always hydrophobic (Fig. 1). Subsequently, we observed that both molecular weight and end-group termination can have order of magnitude effects on the transfection potential of poly(h-amino esters) [10]. Therefore, we sought to clarify further the polymer properties that result in optimal transfection. As previously mentioned, the synthesis of poly(hamino esters) proceeds via the conjugate addition of amines to acrylate groups. This step in polymerization results in a broad, statistical distribution of polymer lengths, with molecular weight and chain end groups dependent on the monomer stoichiometry [21,22]. Maximal polymer molecular weight is achieved when the ratio of monomers is stoichiometrically equivalent. An excess of either amine or diacrylate monomer results in a predominance of amine-or acrylate-terminated polymers, respectively. Our original large library was synthesized on an ~70 mg scale to allow for high-throughput manipulation of reagents [12]. To allow for greater control of stoichiometry, and thereby polymer molecular weight and endgroup termination, we performed polymer synthesis in 1–
FIG. 1. Acrylate and amino monomers used to synthesize poly(h-amino ester) library.
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
427
ARTICLE
2 g scale. Furthermore, monomer concentrations were increased to favor intermolecular addition over intramolecular cyclization, thereby allowing for the generation of higher molecular weight polymers. Previously, substantial variation in transfection potential and molecular weight was observed through variation of stoichiometric ratio of amine monomer to acrylate monomer from 0.6 to 1.4 [10]. Therefore, we synthesized polymers at these ratios by weighing 500 mg of amine monomer into vials followed by acrylate monomer resulting in the appropriate stoichiometric ratio. We first performed polymerizations at 958C in the absence of solvent to maximize monomer concentration. In some cases, polymers became overly hard and insoluble, in which case polymers were resynthesized at 608C and in the presence of an additional 2 ml of DMSO. Using these methods, we synthesized 486 polymers. Polymer molecular weight ranged from 1000 to 61,000 Da, with an average molecular weight of 12,000 Da (Fig. 2, Supplemental Table 1). Theoretically, the maximum molecular weight should occur at stoichiometric equiv-
doi:10.1016/j.ymthe.2004.11.015
alence (amine/diacrylate = 1 mol/mol). However, in some cases, the actual maximum in molecular weight occurs with an excess of diacrylate. This is likely due to side reactions (e.g., photoreaction of acrylates) that result in a minor consumption of the diacrylate monomer. Transfection Results We performed transfection experiments with all polymers at six different polymer/DNA ratios to determine the impact of molecular weight, polymer/DNA ratio, and chain end group on transfection efficiency (Fig. 3, Supplemental Table 2). As a model system, we used the COS-7 cell line and a plasmid encoding the firefly luciferase reporter gene (pCMV-Luc). We performed over 12,000 transfections (data obtained in quadruplicate) with these 486 polymers using high-throughput transfection methods [12]. The data displayed in Fig. 3 show the average transfection efficiency of all polymers, at the optimal polymer/ DNA ratio. One hundred percent of the 25 highest transfection levels we observed were achieved at poly-
FIG. 2. Molecular weight of polymer library. Mean polymer molecular weights organized from best transfecting (left) to worst (right) and amine/acrylate ratio from high (front) to low (back). Molecular weight was measured by GPC relative to polystyrene standards. Red arrows indicate polymers made with DMSO as described under Experimental Procedures.
428
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
doi:10.1016/j.ymthe.2004.11.015
ARTICLE
FIG. 3. Transfection proficiency of the polymer library. COS-7 cells were transfected with polymer/DNA complexes at six polymer/DNA ratios in quadruplicate: 10:1, 20:1, 30:1, 40:1, 60:1, and 100:1. The average transfection level in nanograms of luciferase per well of the optimal polymer/DNA ratio is presented. Polymers are organized from highest transfecting (C32) to lowest transfecting (II32) and amine/acrylate ratio from high (front) to low (back). Red arrows indicate polymers made with DMSO as described under Experimental Procedures.
mer/DNA ratios of 40:1 or higher (Fig. 4, Supplemental Fig. 1). Interestingly, none of the lower molecular weight polymers were able to mediate the highest levels of transfection. In general, only polymers with molecular weights above 10,000 Da were able to transfect efficiently (Fig. 4, Supplemental Fig. 1). Consistent with our previous work, the most effective polymers (C32, JJ28, C28) contain alcohol groups together with hydrophobic acrylates. The overall transfection levels and the monomer composition of the most effective polymers in this library are both higher and different from those identified in our preliminary, high-throughput screen. This illustrates the importance of molecular weight, polymer/ DNA ratio, and chain end-group identity. Early in our synthesis and testing, as well as in our previous work [10], we observed that acrylate-terminated polymers trans-
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
fected cells less effectively than amino-terminated polymers. Therefore, the majority of polymers were synthesized at monomer ratios with an excess of the amino monomer. While it is unclear how amino termination improves transfection, the extra positive charge may assist DNA complexation or some other unknown transfection step. Since all polymers were analyzed at the same time point, it is also possible that differences in the time course of gene expression exist between different polymers. This will be examined in future studies. Optimized transfection with these polymers is significantly higher than that with PEI (polymer/DNA 1/1 w/w) and Lipofectamine 2000, which resulted in expression levels of 6 and 21 ng luciferase/well (data not shown), respectively, under the same conditions. Even under optimized conditions as directed by vendor, these poly-
429
ARTICLE
doi:10.1016/j.ymthe.2004.11.015
FIG. 4. Transfection proficiency relative to polymer molecular weight and polymer/DNA ratio. COS-7 cells were transfected with polymer/DNA complexes at six polymer/ DNA ratios in quadruplicate: 10:1, 20:1, 30:1, 40:1, 60:1, and 100:1. The average transfection level in nanograms of luciferase per well is presented for all ratios as a function of molecular weight.
mers are significantly more effective than Lipofectamine 2000 (32 ng luciferase/well—data not shown). Particle Sizing The primary method of cellular entry by nanoparticles is endocytosis. The ability of polymers to condense DNA into nanometer-scale particles is one requirement for cellular entry. To understand better how the polymer molecular weight and structure affect DNA condensation, we complexed DNA to polymers for most of the top performing polymers in our library and used dynamic light scattering to measure the particle size (Fig. 5, Supplemental Table 3). Polymer/DNA complexes were formed at the polymer/DNA ratios found to yield the highest levels of transfection. As expected, the majority of effective polymers condensed DNA to nanometer-sized particles. Particles made with the top performing polymers (C32, JJ28, and U28) had the smallest size (71, 79, and 82 nm, respectively). These results are consistent with the idea that small particle size is an important factor for attaining maximal transfection efficiency. Particle Surface Charge It has been demonstrated that nonspecific cellular uptake of nanoparticles is enhanced by the presence of a positive charge. To understand better the role surface charge plays in poly(h-amino ester)-mediated transfection, we measured the ~ potential of nanoparticles formed with top performing polymers from our library (Fig. 6, Supplemental Table 4). We found that the vast majority of polymers condensed DNA into positively charged nanoparticles. Furthermore, the best performing polymers
430
were always positively charged, usually with a ~ potential over 10 mV.
DISCUSSION In our high-throughput studies of poly(h-amino ester)mediated transfection, we observed that effective transfection polymers contained alcohols, imidazoles, or secondary diamines together with a hydrophobic diacrylate [12]. Subsequently, we showed that poly(h-amino ester)-mediated transfection was greatly dependent on the molecular weight and terminal end group [10]. Since poly(h-amino esters) are synthesized via conjugate addition, both molecular weight and terminal end group are greatly dependent on the stoichiometric ratio of amine and acrylate monomers during synthesis. Here we have systematically examined the effects of polymer molecular weight, polymer end group, polymer/ DNA ratio, polymer/DNA particle size, particle surface charge, and chemical composition on transfection by poly(h-amino esters). Consistent with previous work with cationic DNA delivery polymers, the most effective poly(h-amino esters) all induce the condensation of DNA into small, positively charged nanoparticles (Figs. 5 and 6 and Supplemental Tables 3 and 4). Interestingly, the two most effective transfection polymers, C32 (1.2/1 amine/acrylate ratio) and JJ28 (1.1/1 amine/acrylate ratio), condensed DNA into the smallest particles (71 and 79 nm, respectively). However, other polymers that were much less effective (for example D32—1.3/1 amine/ acrylate ratio, particle size 91 nm) also condensed DNA to sub-150-nm-sized particles. Therefore, DNA condensation
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
doi:10.1016/j.ymthe.2004.11.015
ARTICLE
FIG. 5. Relationship of polymer/DNA particle size and transfection efficiency. Polymer/DNA complexes were formed at polymer/DNA ratios optimal for transfection (Fig. 3). Data are organized from highest transfecting (left) to lowest transfecting (right) and amine/acrylate ratio from high (front) to low (back). Red arrows indicate polymers made with DMSO as described under Experimental Procedures.
and nanoparticle size alone cannot explain the efficiency of certain polymers over others. Although particle size and stability are important factors impacting cellular uptake, it is likely that other factors such as endosomal escape, DNA protection, and decomplexation, also play a critical role in determining transfection efficiency. By varying monomer ratios, we explored the effects of polymer molecular weight and chain termination groups on transfection efficiency. For a given monomer composition, the effects on transfection could vary tremendously—for example, C32 synthesized with a 1.2/1 amine/acrylate ratio is ~45-fold more effective than C32 synthesized with a 1.025/1 amine/acrylate ratio. This difference in transfection is likely an effect of both the molecular weight differences and the chain terminal differences (Fig. 2, Supplemental Table 1). However, C32 synthesized at 1.05/1 amine/acrylate ratio was ~8fold more effective than C32 synthesized at 1/1 ratio, despite the fact that the average molecular weights of
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
both polymers are the same. This likely reflects the importance of polymer chain termination, as the excess of amine monomer during synthesis results in amineterminated polymer. This suggests that further optimization of polymer transfection efficiency may be possible by simple end modification of poly(h-amino esters) to include amines or other charged or functional groups. As expected from previous work [10], analysis of transfection relative to molecular weight (Fig. 4, Supplemental Fig. 1) reveals that low-molecular-weight polymers were less effective at DNA delivery. Specifically, transfection potential was proportional to polymer molecular weight, with the highest transfection levels achieved only with polymers of average molecular weight over approximately 10,000 Da.. This minimal molecular weight likely relates to the affinity of the polymers to DNA [20]. Interestingly, of the 486 different polymers synthesized, the 9 most effective polymer structures are all
431
ARTICLE
doi:10.1016/j.ymthe.2004.11.015
FIG. 6. Relationship of polymer/DNA complex surface charge and transfection efficiency. Polymer/DNA complexes were formed at polymer/DNA ratios optimal for transfection. ~ potential is presented in mV. Data are organized from highest transfecting (left) to lowest transfecting (right) and amine/acrylate ratio from high (front) to low (back). Red arrows indicate polymers made with DMSO as described under Experimental Procedures.
formed from amino alcohols (Fig. 3, Supplemental Table 2). In fact, the structure of the 3 top performing polymers (C32, JJ28, and C28) differs by only one carbon, being formed from an amino alcohol (monomer 32 or 28), together with a simple diacrylate (C or JJ). While the convergence of polymer structure suggests a common mode of action, the precise method by which these polymers elicit such high levels of transfection remains unclear. Previously, it was shown that some poly(hamino esters) are capable of buffering endosomal compartments [11,23]. It is possible that the structure common to our lead polymers may be particularly effective at inducing endosomal lysis or buffering. Through these approaches we have generated a library of degradable polymer-based vectors that rival the best available nonviral vectors for in vitro gene transfer. Furthermore, we have identified several common features
432
that may provide a basis for the generation of even more effective nonviral DNA delivery systems. (1) Of all 486 polymers, the 3 most effective polymers share a common structure (C32, JJ28, and C28). We are currently attempting to optimize this structure further through further modification of the side chains and the addition of targeting ligands. This avenue has shown to be particularly effective in enhancing the utility of PEI [24–26]. (2) The observation that the smallest nanoparticles also transfect at the highest levels suggests that a further decrease in complex size may increase the efficiency of this and other nonviral delivery systems. It may be possible to reduce particle size further through optimization and decrease of DNA size, as well as further engineering of polymer structure. (3) We have observed that end termination of polymers is critical for optimal function. We are currently modifying the end terminal
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
doi:10.1016/j.ymthe.2004.11.015
regions of these polymers to increase charge density, as well as tissue-specific ligands. Future studies examining the ability of these polymers to transfect different cell types as well as examining in vivo function are currently in progress.
EXPERIMENTAL PROCEDURES Polymer Synthesis Monomers were purchased from Aldrich (Milwaukee, WI, USA), TCI (Portland, OR, USA), Pfaltz & Bauer (Waterbury, CT, USA), Matrix Scientific (Columbia, SC, USA), Scientific Polymer (Ontario, NY, USA), and Dajac Monomer–Polymer (Feasterville, PA, USA). Six to 12 versions of each polymer were generated by varying the amine/ diacrylate stoichiometric ratio. To synthesize each polymer, 400 mg of amino monomer was weighed into an 8ml sample vial with Teflon-lined screw cap. Next, the appropriate amount of diacrylate was added to the vial to yield a stoichiometric ratio ranging from 0.6 to 1.4. A small Teflon-coated stir bar was then put in each vial. Polymers were then synthesized on a multiposition magnetic stir-plate residing in an oven at (1) 958C and solvent free or (2) 608C with 2 ml DMSO added. Hightemperature synthesis was performed for approximately 12 h, and low-temperature synthesis was performed for 2 days. After completion of reaction, all vials were removed from the oven and stored at 48C. All polymers were analyzed by gel-permeation chromatography (GPC). Gel Permeation Chromatography GPC was performed using a Hewlett–Packard 1100 Series isocratic pump, a Rheodyne Model 7125 injector with a 100-Al injection loop, and a Phenogel MXL column (5 Am mixed, 300 7.5 mm; Phenomenex, Torrance, CA, USA). Seventy percent THF/30% DMSO (v/v) + 0.1 M piperidine was used as the eluent at a flow rate of 1.0 ml/min. Data were collected using an Optilab DSP interferometric refractometer (Wyatt Technology, Santa Barbara, CA, USA) and processed using the TriSEC GPC software package (Viscotek Corp., Houston, TX, USA). The molecular weights of the polymers were determined relative to monodisperse polystyrene standards. Luciferase Transfection Assays COS-7 cells (ATCC, Manassas, VA, USA) were seeded (14,000 cells/well) into each well of an opaque white 96well plate (Corning-Costar, Kennebunk, ME, USA) and allowed to attach overnight in growth medium. Growth medium was composed of 90% phenol red-free DMEM, 10% fetal bovine serum, 100 units/ml penicillin, 100 Ag/ ml streptomycin (Invitrogen, Carlsbad, CA, USA). To facilitate handling, polymer stock solutions (100 mg/ml) were prepared in DMSO solvent. (Note. We have demonstrated that the small residual amount of DMSO in the transfection mixture does not affect transfection effi-
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy
ARTICLE
ciency and does not result in any observable cytotoxicity.) Working dilutions of each polymer were prepared (at concentrations necessary to yield the different polymer/DNA weight ratios) in 25 mM sodium acetate buffer (pH 5). Twenty-five microliters of the diluted polymer was added to 25 Al of 60 Ag/ml pCMV-Luc DNA (Elim Biopharmaceuticals, South San Francisco, CA, USA) in a well of a 96-well plate. The mixtures were incubated for 10 min to allow for complex formation, and then 30 Al of each of the polymer/DNA solutions was added to 200 Al of Opti-MEM with sodium bicarbonate (Invitrogen) in 96-well polystyrene plates. The growth medium was removed from the cells using a 12-channel aspirating wand (V&P Scientific, San Diego, CA, USA) after which 150 Al of the Opti-MEM/polymer/DNA solution was immediately added. Complexes were incubated with the cells for 1 h and then removed using the 12-channel wand and replaced with 105 Al of growth medium. Cells were allowed to grow for 3 days at 378C, 5% CO2 and were then analyzed for luciferase expression. Control experiments were also performed with PEI (MW 25,000; Sigma–Aldrich) and Lipofectamine 2000 (Invitrogen). PEI transfections were performed as described above, but using polymer/DNA weight ratios of 1/1. Lipofectamine 2000 transfections were performed as described by the vendor, except that complexes were removed after 1 h in some cases. Luciferase expression was analyzed using Bright-Glo assay kits (Promega, Madison, WI, USA). Briefly, 100 Al of Bright-Glo solution was added to each well of the 96well plate containing medium and cells. Luminescence was measured using a Mithras luminometer (Berthold, Oak Ridge, TN, USA). A 1% neutral density filter (Chroma, Brattleboro, VT, USA) was used to prevent saturation of the luminometer. A standard curve for luciferase was generated by titration of luciferase enzyme (Promega) into growth medium in an opaque white 96well plate. Measurement of Particle Size and Z Potential Particle size and ~ potential measurements were made using a ZetaPALS dynamic light scattering detector (Brookhaven Instruments Corp., Holtsville, NY, USA; 15-mW laser, incident beam 676 nm). Polymer/DNA complexes were prepared as described above at the polymer/DNA ratio determined to have the highest transfection efficiency for each particular polymer. Complexes were then diluted in 1.4 ml of 25 mM Hepes buffer, pH 7.2. Correlation functions were collected at a scattering angle of 908, and particle sizes were calculated using the MAS option of BICTs particle sizing software (version 2.30) using the viscosity and refractive index of pure water at 258C. Particle sizes are expressed as effective diameters assuming a log-normal distribution. Average electrophoretic mobilities were measured at 258C using BIC PALS ~ potential analysis software, and ~ potentials
433
ARTICLE
were calculated using the Smoluchowsky model for aqueous suspensions. ACKNOWLEDGMENT We thank Janet Sawicki and Steve Little for helpful comments. This work was supported by the NSF (through the MIT Biotechnology Process and Engineering Center) and NIH Grant EB 00244.
APPENDIX A. SUPPLEMENTARY DATA Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ymthe. 2004.11.015. RECEIVED FOR PUBLICATION SEPTEMBER 28, 2004; ACCEPTED NOVEMBER 22, 2004.
REFERENCES 1. Journal of Gene Medicine clinical trial site (2002). http://www.wiley.co.uk/ genetherapy/clinical/. 2. Ledley, F. D. (1995). Nonviral gene-therapy—the promise of genes as pharmaceutical products. Hum. Gene Ther. 6: 1129 – 1144. 3. Luo, D., and Saltzman, W. M. (2000). Synthetic DNA delivery systems. Nat. Biotechnol. 18: 33 – 37. 4. Fominaya, J., and Wels, W. (1996). Target cell-specific DNA transfer mediated by a chimeric multidomain protein—novel non-viral gene delivery system. J. Biol. Chem. 271: 10560 – 10568. 5. Schwartz, J. J., and Zhang, S. (2000). Peptide-mediated cellular delivery. Curr. Opin. Mol. Ther. 2: 162 – 167. 6. Boussif, O., et al. (1995). A versatile vector for gene and oligonucleotide transfer into cells in culture and in-vivo—polyethylenimine. Proc. Natl. Acad. Sci. USA 92: 7297 – 7301. 7. Kabanov, A. V., Gebhart, C. L., Bronich, T. K., and Vinogradov, S. V. (2002). Polycations for gene delivery: problems and solutions. Abstr. Am. Chem. Soc. 224: U471. 8. Lim, Y. B., Kim, C. H., Kim, K., Kim, S. W., and Park, J. S. (2000). Development of a safe gene delivery system using biodegradable polymer, poly[alpha-(4-aminobutyl)-Lglycolic acid]. J. Am. Chem. Soc. 122: 6524 – 6525. 9. Lim, Y. B., Kim, S. M., Suh, H., and Park, J. S. (2002). Biodegradable, endosome disruptive, and cationic network-type polymer as a highly efficient and nontoxic gene delivery carrier. Bioconjugate Chem. 13: 952 – 957.
434
doi:10.1016/j.ymthe.2004.11.015
10. Akinc, A., Anderson, D. G., Lynn, D. M., and Langer, R. (2003). Synthesis of poly(betaamino ester)s optimized for highly effective gene delivery. Bioconjugate Chem. 14: 979 – 988. 11. Akinc, A., Lynn, D. M., Anderson, D. G., and Langer, R. (2003). Parallel synthesis and biophysical characterization of a degradable polymer library for gene delivery. J. Am. Chem. Soc. 125: 5316 – 5323. 12. Anderson, D. G., et al. (2003). Semi-automated synthesis and screening of a large library of degradable cationic polymers for gene delivery. Angewandte Chemie-Int. Ed. 42: 3153 – 3158. 13. Lynn, D. M., Anderson, D. G., Putnam, D., and Langer, R. (2001). Accelerated discovery of synthetic transfection vectors: parallel synthesis and screening of degradable polymer library. J. Am. Chem. Soc. 123: 8155 – 8156. 14. Lynn, D. M., and Langer, R. (2000). Degradable poly(beta-amino esters): synthesis, characterization, and self-assembly with plasmid DNA. J. Am. Chem. Soc. 122: 10761 – 10768. 15. Brocchini, S., James, K., Tangpasuthadol, V., and Kohn, J. (1997). A combinatorial approach for polymer design. J. Am. Chem. Soc. 119: 4553 – 4554. 16. Anderson, D. G., Levenberg, S., and Langer, R. (2004). Nanoliter-scale synthesis of arrayed biomaterials and application to human embryonic stem cells. Nat. Biotechnol. 22: 863 – 866. 17. Murphy, J. E., Uno, T., Hamer, J. D., Cohen, F. E., Dwarki, V., and Zuckermann, R. N. (1998). A combinatorial approach to the discovery of efficient cationic peptoid reagents for gene delivery. Proc. Natl. Acad. Sci. USA 95: 1517 – 1522. 18. Godbey, W. T., Wu, K. K., and Mikos, A. G. (1999). Size matters: molecular weight affects the efficiency of poly(ethylenimine) as a gene delivery vehicle. J. Biomed. Mater. Res. 45: 268 – 275. 19. Kabanov, A. V., Vinogradov, S. V., Suzdaltseva, Y. G., and Alakhov, V. (1995). Watersoluble block polycations as carriers for oligonucleotide delivery. Bioconjugate Chem. 6: 639 – 643. 20. Schaffer, D. V., Fidelman, N. A., Dan, N., and Lauffenburger, D. A. (2000). Vector unpacking as a potential barrier for receptor-mediated polyplex gene delivery. Biotechnol. Bioeng. 67: 598 – 606. 21. Flory, P. J. (1953). Principles of Polymer Chemistry. Cornell Univ. Press, Ithaca, NY. 1953. 22. Odian, G. G. (1991). Principles of Polymerization. Wiley, New York. 23. Akinc, A., and Langer, R. (2002). Measuring the pH environment of DNA delivered using nonviral vectors: implications for lysosomal trafficking. Biotechnol. Bioeng. 78: 503 – 508. 24. Thomas, M., and Klibanov, A. M. (2002). Enhancing polyethylenimineTs delivery of plasmid DNA into mammalian cells. Proc. Natl. Acad. Sci. USA 99: 14640 – 14645. 25. Kunath, K., Merdan, T., Hegener, O., Haberlein, H., and Kissel, T. (2003). Integrin targeting using RGD-PEI conjugates for in vitro gene transfer. J. Gene Med. 5: 588 – 599. 26. Kunath, K., von Harpe, A., Fischer, D., and Kissel, A. (2003). Galactose–PEI–DNA complexes for targeted gene delivery: degree of substitution affects complex size and transfection efficiency. J. Controlled Release 88: 159 – 172.
MOLECULAR THERAPY Vol. 11, No. 3, March 2005 Copyright C The American Society of Gene Therapy