Steric block oligomers such as of 2'O-methyl (2'OMe) and phosphorodiamidate morpholino (PMO) oligonucleotides possess high affinity for their complementary pre-mRNA targets and can modulate alternative splicing, correct aberrant splicing, and induce skipping or inclusion of specific exons (for reviews see 12). These splice modulating oligonucleotides (SMOs) represent a powerful class of compounds with broad utility for basic and translational research and are poised to show rapid growth as pharmaceuticals. However, for many in vivo applications, SMOs administered alone show an inadequate delivery profile for reaching target cell nuclei, necessitating the use of carriers. Indeed, inefficient delivery of SMOs remains the foremost limitation to their usefulness as pharmaceuticals.

Duchenne muscular dystrophy (DMD) is a fatal x-linked disease caused by mutations in the gene encoding the 427 kDa membrane-associated cytoskeletal protein dystrophin, resulting in progressive body-wide muscle weakening and death usually in the early to mid third decade of life. DMD mutations are most often comprised of insertions and deletions that alter the dystrophin reading frame or encode premature stop codons 3. These types of mutations result in production of truncated and non-functional dystrophin that is rapidly degraded.

Steric block exon skipping oligonucleotides (ESOs) have been shown in cultured mouse, canine, and human cells to cause skipping of targeted dystrophin exons, resulting in production of full-length dystrophin mRNA (minus only the skipped exons), and restoration of the reading frame 45678910. The classical animal model for DMD, the mdx mouse, has a point mutation in dystrophin exon 23 that produces a premature stop codon. ESOs have been shown to promote skipping of exon 23 and concomitant dystrophin expression in skeletal muscles of mdx mice after both local and systemic delivery 811121314151617, and this technology in the context of therapy for DMD has been recently reviewed 1819. However, the efficiency of 2'OMe ESO delivery, and level of concomitant dystrophin expression in the mdx mouse remains relatively modest 111316. Despite the fact that ESOs without a carrier have progressed to phase I clinical trials, improved carriers must be developed before the ESO approach can be considered as a viable therapeutic for improving health and longevity in DMD patients.

The amine-rich cationic polymer poly(ethylene imine) (PEI) is a well-studied compound that is effective at condensing large plasmid DNA, enabling improved cellular uptake 202122232425. Although most often applied to plasmid delivery, recent studies have also documented PEI-enhanced delivery of small nucleic acid agents such as oligonucleotides and siRNA 2627282930313233. The positive surface charge of PEI-nucleic acid polyplexes interacts with negatively charged elements on the cell membrane, stimulating non-specific receptor-mediated endocytotic uptake 202124253435. Once internalized, the 'proton sponge effect' enabled by PEI's buffering capacity induces the rupturing of the endosomal compartment due to osmotic lysis 34363738. While the intracellular dynamics of PEI-nucleotide complexes once released from the endosome are unknown, oligonucleotides must dissociate from PEI to reach the nucleus, as the nuclear envelope is likely impermeable to PEI. Using dual-fluorescence tracking we recently showed PEG-PEI copolymers enhanced SMO delivery to myonuclei of cultured mdx myofibers, while the copolymers were mainly excluded from entering the nuclei 39.

The functionality and biocompatibility of PEI is greatly improved by incorporation of the nonionic linear polymer poly(ethylene glycol) (PEG) into PEG-PEI copolymers 2324. The macromolecular properties of PEG-PEI-oligonucleotide polyplexes are greatly influenced by the molecular weight of PEI and nature of PEGylation, which in-turn effects transfection capacity 222340. We recently showed that PEG-PEI copolymers made of low MW PEI2K significantly improved delivery of ESOs to myofibers of mdx mice after intramuscular injections compared to high MW PEI25K copolymers 13. We attributed the superior efficacy of the PEI2K-based copolymers to the low surface charge and high stability of the nanoparticles formed during complexation with ESOs 41. Although intramuscular injection of these PEG-PEI-ESO polyplexes produced significantly more dystrophin-positive fibers than ESO alone, the level of dystrophin expression by Western blot reached only 2–5% of the normal level in TA muscles. Therefore, in this report we evaluated whether new injection regimens and further functionalization of PEG-PEI copolymers with gold nanoparticles might improve dystrophin expression. Our results show that repeat injections of small amounts of PEG-PEI-ESO are very effective at transducing dystrophin expression, reaching up to 20% of normal levels under the most favorable formulation and injection regimen.

The growing opinion that ESOs are on the path to becoming a viable therapeutic option for DMD is well supported by cell culture and animal data showing specific skipping of various targeted exons and resultant induction of nearly full-length dystrophin 46791112131415164950. However, there is vigorous debate as to which ESO chemistry may work best, and what type of carrier compound will provide adequate delivery to body musculature. ESOs of phosphorothioate 2'OMe and PMO chemistry have been the best studied in animal models for DMD and both are currently being tested in Phase I clinical safety trials 51. Both 2'OMe and PMO ESOs function by sterically blocking pre-mRNA target sequences, and it is thought that they may be used interchangeably once an optimal sequence (for a single chemistry) has been empirically determined 52. However, fundamental differences in 2'OMe and PMO backbone chemistry preclude the use of a universal carrier for efficient delivery. Specifically, PMOs are synthetic compounds that are extraordinarily resistant to chemical degradation, but they are also charge neutral which limits cell surface interactions and cellular uptake. The non-degradable nature of PMOs also raises concerns over their safety after extended applications In contrast, 2'OMe ESOs are anionic RNA, and despite improved stability due to their phosphorothioate backbone, remain somewhat susceptible to degradation, while the negative charge hinders biodistribution and cellular uptake.

In this study we showed that PEG-PEI copolymers formulated with low MW PEI2K function as effective carriers for delivery of 2'OMe ESOs to myofibers of mdx mice after intramuscular injections, resulting in improved levels of dystrophin expression. Specifically, three weekly intramuscular injections of only 5 μg of ESO complexed with the PEI2K(PEG550)₁₀ copolymer resulted in about 600 dystrophin-positive fibers and about 11% of the normal level of dystrophin expression at 3 weeks after the initial injection. Still higher levels of dystrophin expression were achieved using a twice-weekly injection regimen extended out to 6 weeks. Specifically, 10 consecutive injections of the NG-PEI2K(PEG550)₁₀ copolymer complexed with 5 μg of ESO produced over 1200 dystrophin positive fibers and 20% of normal levels of dystrophin expression. In regions with the most highly transfected fibers, we observed a concomitant increase in membrane-associated nNOS, specifically in dystrophin-positive fibers, in agreement with previous reports using this ESO 11. Our lab, as well as other, have demonstrated that intramuscular injections of 2'OMe ESO alone, using very similar conditions as used herein, produced very few dystrophin-positive fibers and negligible levels of dystrophin on western blots 1343. In addition, we previously showed that single injections of the PEI2K(PEG550)₁₀ copolymer complexed with 20 μg of ESO produced about 460 dystrophin-positive fibers at 3 weeks after the injection, but western blots showed dystrophin expression was only 2–5% of normal levels 13. Taken together, the current results suggest that PEG-PEI copolymers enhance dystrophin expression and that repeat injections are more effective at transfecting a greater number of muscle fibers than individual injections containing about the same amount of ESO.

Although the dystrophin expression levels shown in this report demonstrate the utility of PEG-PEI nanopolymers for delivery of ESOs, these compounds appear to be somewhat less effective than PMOs. Specifically, Alter et al 14 recently showed that single intramuscular injections of 10 μg of PMO resulted in up to 60% of the normal level of dystrophin expression, although this report appeared to provide only an estimate of efficacy and lacked statistical validation. Other recent studies of PMOs with conjugated peptide cell-targeting moieties showed impressive numbers of dystrophin-positive fibers, but did not provide a thorough evaluation of dystrophin expression by western blots 12.

The PEI2K(PEG550)₁₀ and PEI2K(PEG5K)₁₀ copolymers utilized in this study were previously shown to form exceptionally stable complexes when mixed with negatively charged ESO and the surface charge of the resultant nanoparticulates was relatively low 41. We propose that the high stability and low surface charge of these polyplexes are two salient features that make them better suited for in vivo delivery of ESOs than high MW PEI25K-based copolymers. Specifically, the low polyplex surface charge favors biodistribution and reduces cytotoxicity, while the high stability allows the polyplex to remain associated during extracellular to intracellular trafficking.

Gold nanoparticles such as NG and CG have been shown to improve biocompatibility and enhance cellular uptake of various types of cargo in drug delivery applications 44455354. In particular, NG conjugated to low MW PEI2K showed at least an order of magnitude greater efficiency than PEI25K and was 12 times more potent than unmodified PEI2K for delivery of plasmid DNA in cell culture 44. Unexpectedly, our results showed that neither covalent conjugation of NG or electrostatic surface coating with CG of the PEG-PEI copolymers improved ESO delivery. Polyplex stability assays in PBS showed CG and NG caused only moderate weakening of polyplex stability (data not shown), making this an unlikely explanation for the lack of improved delivery. A possible explanation for the lack of improvement is that the CG coating was not stable enough to adhere to the copolymer during delivery. On the other hand, NG was covalently conjugated to PEG-PEI, and we postulate that its ineffectiveness was more likely due to the 1:10 NG to PEI2K ratio, which may have been too low to improve functionality.

The dystrophin expression achieved in the present study was accomplished with PEG-PEI carriers that did not appear to elicit any overt signs of cytotoxicity. This is in contrast to previous studies using non-PEGylated PEI25K as a carrier of 2'OMe ESOs, which was ineffective and resulted in significant damage following only a very limited number of injections 5556. Based on these results, it was concluded that cationic polymers are unsuitable for in vivo delivery of AO in skeletal muscle 1157. However, the PEI-nucleotide particles used in these previous studies undoubtedly had very high positive surface charges, because they did not contain PEG, which is known to provide steric shielding of the PEI surface charge. In addition, dispersion of the highly-charged particles after intramuscular injection is probably severely hindered by charge interactions between the PEI and negatively-charged elements within the extracellular environment. Therefore, it is not surprising that these previous muscle transfection studies with non-PEGylated PEI produced unsatisfactory results. We suggest that the combination of a low MW PEI and extensive PEGylation used presently provided a favorable formulation which was both effective and non-toxic. However, the lack of cytotoxicity observed does not preclude the possibility that some damage to muscle occurs immediately after injection, resulting in some level of degeneration-regeneration. This process may underlie to some extent the high number of dystrophin-positive fibers observed in our 6 week (10 injection) trials. Although not systematically evaluated, we have observed that short-term mechanical damage occurs in mdx muscles after intramuscular injections of various solutions (even saline) that does not occur in normal muscle. Because of the lack of dystrophin, mdx muscles are more susceptible to mechanical damage than normal muscle. This effect may be exacerbated to some extent by cationic particles, or for that matter, any type of carrier compound.

We suggest that the major limitation of the carrier-ESO formulations described in this report was inadequate carrier functionality, and not a lack of intrinsic potency of the ESO. The ESO used in the present study (designated in the literature as M23D(+02–18)) has been shown in vitro to predominantly produce skipping of exon 23, although some exon 22–23 double skipping does occur 42. Moreover, in this study we showed that under the most effective condition, about 50% of fibers were dystrophin-positive, resulting in about 20% of normal dystrophin expression. This indicates that on average dystrophin-positive fibers contained about 40% of the normal level of dystrophin. A similar calculation based on data reported by Lu et al 11, and our recent study with TAT-conjugated copolymers (Sirsi et al., manuscript submitted) suggests that dystrophin per transfected fiber may reach 75% of normal levels. Thus, the main limitation with cationic carriers seems to be their poor diffusional distribution, as indicated by large regions in muscles with no apparent transfection. Thus, further improvements in carrier functionality will likely be required to enable their usage in a clinical setting for DMD. Our group recently showed that conjugation of multiple HIV-TAT epitopes to PEI2K(PEG5K)₁₀ copolymers greatly improved ESO delivery, using a similar dosing and intramuscular injection regimen as reported here, resulting in up to 30% dystrophin expression (Sirsi et al., manuscript submitted). Various other types of cell targeting ligands, cell penetrating peptides, or fusogenic peptides may also be conjugated to PEI to improve functionality. Importantly, this type of peptide-PEI conjugate can likely be formulated for improved systemic delivery, which will be required to achieve meaningful therapeutic benefit. PMOs have already been shown to have limited efficacy after systemic delivery. For example, intraperitoneal injections into neonatal mdx mice of 5 mg/kg/week with PMO-peptide produced widespread dystrophin in diaphragm muscle, with low levels of expression observed in limb muscles 1258.

In this report, we show that PEGylated amine-rich branched cationic nanopolymers comprised of low MW PEI2K are effective carriers for delivery of ESOs to myofibers of mdx mice after intramuscular injections. Our results indicate that high frequency, low-dosage, long-term injection regimens using these carrier-ESO compounds provide the most favorable outcome in terms of dystrophin expression. While other studies have shown improvements utilizing gold nanoparticles, we showed no enhancement when conjugated to the current PEG-PEI formulations. Despite the lack of improvement using these novel conjugations, PEGylated PEI2K copolymers remain as one of the most efficient carriers for local delivery of 2'OMe ESOs and warrant further development as potential therapeutics for treatment of DMD.

ESO, Exon skipping oligonucleotide; 2'OMe, 2'-O-Methyl; PMO, Phosphorodiamidate Morpholino Oligonucleotide; CPP, Cell Penetrating Peptide; DMD, Duchenne Muscular Dystrophy; PEI, Poly(ethylene imine); PEG, Poly(ethylene glycol); N:P, ratio of PEI nitrogen to ESO phosphate; TA, Tibialis Anterior; NG, nanogold; CG, Colloidal gold.

The author(s) declare that they have no competing interests.

JHW carried out western and immunoflourescence analysis, assisted with harvesting of muscle tissue, participated in the design of the study and drafted the manuscript. RCS maintained animal colonies, carried out injections and harvesting of tissues and participated in design of the study. SRS carried out chemical synthesis and functionalization of copolymers and contributed to the conception of the study. GJL conceived the study, drafted the final manuscript and participated in all stages of the work. All authors read and approved the final manuscript.