Gene Therapy (2003) 10, 131–142 & 2003 Nature Publishing Group All rights reserved 0969-7128/03 $25.00 www.nature.com/gt RESEARCH ARTICLE Non-viral gene delivery in skeletal muscle: a protein factory

QL Lu1, G Bou-Gharios1 and TA Partridge1 1Muscle Cell Biology Group, MRC Clinical Sciences Centre, Faculty of Medicine, Imperial College of Science, Technology and Medicine, London, UK

Ever since the publication of the first reports in 1990 using and hormones. Although the most efficient way of introducing skeletal muscle as a direct target for expressing foreign transgenes into muscle fibres has been by a variety of transgenes, an avalanche of papers has identified a variety recombinant viral vectors, there are potential benefits in the of proteins that can be synthesized and correctly processed use of non-viral vectors. In this review we assess the recent by skeletal muscle. The impetus to the development of such advances in construction and delivery of naked plasmid DNA applications is not only amelioration of muscle diseases, but to skeletal muscle and highlight the options available for also a range of therapeutic applications, from immunization further improvements to raise efficiency to therapeutic levels. to delivery of therapeutic proteins, such as clotting factors Gene Therapy (2003) 10, 131–142. doi:10.1038/sj.gt.3301874

Keywords: plasmid DNA; skeletal muscle; gene delivery

Introduction plasmid DNA itself has low immunogenicity and can be administered repeatedly although local inflammatory Transgenes can be delivered to muscle using viral or reactions are still a concern.19 The other major disadvan- non-viral systems. Although the pioneer study used tage reported is that plasmid DNA-mediated gene naked plasmid DNA for gene delivery,1 the low delivery with a foreign transgene driven by non-muscle efficiency of this approach has led to the focus being specific promoters may evoke more severe immune placed on viral delivery systems over the last 10 years or responses to the transgene product than AAV vectors so. Several genetically modified viruses, such as retro- expressing the same protein.20,21 virus, herpes simplex virus, Epstein–Barr virus, adeno- The big drawback of plasmid DNA as a vector virus (AdV) and adeno-associated virus (AAV), have all is its low efficiency for gene delivery. However, this been tested for gene delivery into muscle.2–5 Of these, has recently been improved significantly by use of AdV and AAV have been found to be most efficient so far transfection-enhancing reagents such as lipids and for transducing muscle fibres.6–8 AAV in particular is able polymers. These reagents improve the dissemination to transfect non-proliferating muscle fibres and is of plasmids within the target tissues, penetration considered to be non-pathogenic, as it is not found to through cell membranes, stability of plasmids within be associated with any disease. Gene delivery by AAV the target cells, and entry of plasmids into nuclei. In vector also appears to be the least immunogenic of the addition, novel methods have been developed for viral vector systems as regards the transgene and can restricted local as well as systemic delivery of transgenes, sustain longer periods of transgene expression than other including the use of microbubbles and ultrasound, multi- viral vectors.9–11 While the majority of gene therapy purpose gene delivery injectors, and, perhaps the experiments and clinical trials currently use viral most promising, electroporation. Another development delivery systems,12–14 several aspects need to be im- in non-viral delivery is the use of viral-specific peptide proved to achieve therapeutic benefit in patients. These sequences or synthetic peptides, aimed at obtaining include better efficiency and reliability of procedures for high efficiency and tissue specificity without the the manufacture of virus, freedom from helper virus use of infectious virus or whole viral coat proteins. contamination,15 avoidance of immunogenicity of the These developments in combination with advances in viral particle itself, and ways of overcoming the limita- vector construction have led to efficiencies of transgene tion of transgene size.16–18 expression with non-viral delivery systems comparable In several of these respects, non-viral delivery systems to those with recombinant viral vectors.22 In this review, have clear advantages over viral delivery systems. The we will focus on these developments in non-viral use of plasmid DNA does not pose health risks in delivery systems aimed at using skeletal muscle as a principle entailed by viral infection. Plasmid DNA is target tissue. easy to propagate on a large scale at high quality, and is also able to carry relatively large DNA sequences. Naked Why skeletal muscle?

Correspondence: Dr QL Lu, Muscle Cell Biology Group, MRC Clinical Skeletal muscle, by virtue of a number of inherent Sciences Centre, Faculty of Medicine, Imperial College of Science, anatomical, cellular and physiological properties, com- Technology and Medicine, Hammersmith Campus, London W12 0NN, UK mends itself as a target tissue for gene therapy, Non-viral gene delivery in skeletal muscle QL Lu et al 132 particularly for production of proteins as systemic shown little beneficial effect when used for in vivo gene therapeutic reagents. transfer into skeletal muscle (see below).

(1) Skeletal muscle constitutes about 30% of the normal adult body mass and is easily accessible for nearly Vector construction all gene delivery approaches currently used in gene therapy. The importance of vector construction in gene therapy is unequivocal for both viral and naked plasmid delivery (2) Skeletal muscle has an abundant blood vascular systems. Vectors can influence the level and duration of supply. An extensive capillary network wraps gene expression, affect the efficiency of gene delivery around every muscle fibre with regular spacing, and confer regulatory functions. thus providing an efficient transport system for the carriage of secreted protein into the circulation. Promoter, enhancer and other elements (3) Skeletal muscle fibres are terminally differentiated The most critical feature of a vector is the ‘promoter’ cells, and nuclei within the fibres are post-mitotic. which drives the transcription of a transgene. A variety Individual fibres are thought to persist for much of of plasmid backbones containing different promoters the lifetime of the individual. Even when muscle have been tested for transgene expression in skeletal fibres are damaged, only short segments of muscle with varying success. The most widely used are individual fibres undergo degeneration, and the viral promoters, such as cytomegalovirus (CMV) im- myonuclei of surviving segments remain viable.23 mediate-early promoter,28 simian virus, SV40 early This provides a stable environment as a ‘‘factory’’ promoter29 and Rous sarcoma virus promoter.1 These for continuous production of transgene. promoters can drive expression in a wide range of cells (4) The syncytial nature of muscle fibres provides a and tissues eg CMV promoter can be activated in nearly mechanism for dispersal of transgene from a limited all cell types. Of the non-viral promoters employed, some such as b actin have general activity within site of penetration to a large number of 30,31 neighbouring nuclei within the fibre. Such eukaryote cells. Some promoters control expression dispersion within transfected fibres may well be in a cell- or tissue-specific manner eg a skeletal muscle actin promoter,32 myosin light chain 3F promoter33 and one of the reasons behind the efficient expression of 34 transgenes in muscle. muscle creatine kinase (MCK) promoter which are activated specifically in muscle. The best example of Skeletal muscle also has great regenerative capacity promoters of high cell specificity is the ventricle-specific due to the presence of a population of stem-like myosin light chain-2 which is activated exclusively in precursor cells, called satellite cells, situated between cardiac myocytes.35 fibre membrane and basal lamina. These cells, while Muscle-specific promoters may have advantages over quiescent in normal muscle, can be activated in response non-muscle-specific promoters, particularly for gene to muscle damage, proliferating, migrating and fusing therapy of muscular dystrophies. Cordier et al, using an with one another to replace lost fibres, resulting in a AAV vector for the gamma-sarcoglycan construct, regeneration of the lesion (see review by Partridge24). observed a much diminished humoral immune response Moreover, satellite cells can be purified from muscles and a more persistent expression when the transgene and cultured without losing their ability to differentiate was under the control of muscle-specific MCK promoter into muscle fibres (Figure 1). A large number of than when it was controlled by the CMV promoter.36 myogenic cells can be produced from a single muscle They suggested that it may be important to restrict biopsy and transferred back into the host after ex vivo transgene expression to muscle tissue, so that antigen- genetic manipulation (Figure 1), thus providing the presenting dendritic cells are unable to elicit an immune potential to expand and bank an individual’s own cells response. This may be crucial for treatment of muscular expressing any desired transgene. In this context, the dystrophy which requires the long-term expression of recent identification of other stem cell populations within the transgene. In contrast, Wang et al reported that skeletal muscle is also potentially important to gene equally sustained expressions of human mini-dystrophin therapy.25,26 These stem cells, although their precise genes were achieved under the control of both CMV and nature is not understood, are capable of differentiating MCK promoters in otherwise the same AAV-mediated into other cell lineages, thus providing an accessible gene transfer vectors.37 In this case, it may be that the source of multi-potential stem cells for gene therapy. highly deleted human dystophin proteins, lacking al- Having highlighted the advantages of skeletal muscle most the entire rod domain, contained few epitopes that as a target for gene therapy, one needs to bear in mind its were not present on murine dystrophin expressed in possible limitations. For example, post-transcriptional revertant fibres,38 and might thus benefit from immune modifications in muscle may differ from other cell types. tolerance to these proteins. Clearly, we need a better Arruda et al found that carboxylation, tyrosine sulpha- understanding of the interaction of the immune system tion and serine phosphorylation of factor IX differ and transgene products and the influence of various between liver and muscle, although post-translational vector and promoter systems. modifications critical for biological activity of the factor Natural muscle-specific promoters in general drive are similar in these two tissues.27 It should also be noted relatively lower levels of transgene expression compared that there are substantial differences between the to non-tissue-specific viral promoters, such as CMV efficacies of transfection of myoblasts in vitro and muscle promoter, thus limiting their potential use in muscle as a fibres in vivo for any given procedure. Nearly all systems gene therapy target. Efforts have been made to increase that give efficient plasmid delivery in vitro have so far the strength of muscle-specific promoters while

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 133

Figure 1 (a) An isolated muscle fibre showing a chimeric expression of nuclear-localizing b-galactosidase in myonuclei (blue staining and indicated by arrow heads) over half the length (left side) of the fibre. Host myonuclei are stained pink with propidium iodide (indicated by arrows). Donor H2K myoblasts were first infected with retrovirus carrying b-galactosidase transgene and the infected cells were then injected into the tibialis anterior muscle. Muscles were removed 4 weeks later and single fibres were prepared and stained for b-galactosidase and counterstained with propidium iodide. The blue X-gal nuclear staining quenches the propidium iodide fluorescence. (b) Marker gene, GFP, expression in cultured myotubes and myoblasts. H2K myoblasts were cultured and transfected with plasmid DNA containing the GFP transgene mediated by polyethylenimine 25000 (Sigma). Both myoblasts (indicated by arrow heads) and myofibres (indicated by arrows and multi-nucleation) show expression of GFP. GFP signals were seen throughout the whole length of the myofibres. The nuclei are counterstained blue with DAPI.

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 134 maintaining their tissue specificity. Muscle-specific pro- moters display complex organization usually involving combinations of several myogenic regulatory elements including E-box, MEF-2, TEF-1 and SRE sites. Li et al assembled these elements randomly into synthetic promoter recombinant libraries and screened hundreds of individual clones for transcriptional activity both in vitro and in vivo.30 Several artificial promoters were isolated whose transcriptional potencies exceed those of natural myogenic and viral gene promoters. Promoters are not the only important elements in control of transgene expression. Attention must also be paid to enhancers, poly A tails and other regulatory elements such as intronic sequences. The power of a vector to drive gene expression is a combined effect of all these elements and may vary between particular cell or tissue systems. Efforts have been made to construct highly efficient vector systems for transgene expression in muscle by searching for optimal combinations of these elements. Barnhart et al used enhancers and promoters Figure 2 Diagrammatic representation of idealized vector for high levels of from various muscle-specific genes and created 19 transgene expression in skeletal muscle. Note the presence of two enhancer/promoter chimeras in combination with the enhancers, one in front of promoter sequence and the other at the end of human CMV IE enhancer/promoter to drive a luciferase poly A tail of the transgene. The additional nuclear localization element reporter gene.39 The efficiency of reporter gene expres- (NLE) and other regulatory elements would contribute to high level and sion was examined both in vitro by transfection of regulated production of the transgenic protein. differentiated C2C12 mouse myoblasts and in vivo by direct intramuscular injection. Only a 170 bp myogenin enhancer sequence inserted in front of CMV promoter tems.41 Although the mechanism(s) by which plasmid increased expression levels in vivo compared with the DNA enters nuclei is not well established, efforts have CMV promoter alone. In other experiments, Li et al been made to facilitate the nuclear import of plasmids by combined the CMV promoter/enhancer with the SV40 utilizing mechanisms employed by viruses or nuclear enhancer and found a 20-fold greater expression in proteins. Since some viral and non-viral DNA sequences muscle with a single copy of the 72 bp element of the can be recognized by nuclear import machinery, incor- SV40 enhancer placed either 50 of the CMV promoter/ poration of such specific DNA sequences into vectors enhancer or 30 of the poly A tail.40 However, this higher may aid the entry of plasmid DNA into nuclei. For expression was only seen in non-dividing muscle fibres, example, the yeast transcription factor GAL4 enhances not in myoblasts. Similarly, Xu et al evaluated the gene expression through its binding to specific 17 bp combined effect of different promoters, enhancers and DNA sequence and its nuclear localization ability. By other regulatory elements (including promoters and incorporating this DNA sequence into a vector together introns of RSV, human phosphoglycerate kinase and b with the GAL4 protein, Chan et al found an increase in actin), using luciferase as a reporter gene.31 The most transgene expression, which they attributed to GAL4- powerful vector system for gene expression in skeletal mediated transport of the plasmid into the nuclei.42 muscle was a combination of CMV or b actin promoters Similarly, Dean and colleagues showed that a 72-bp with SV40 enhancer, together with SV40 poly A and repeat of the SV40 enhancer facilitated nuclear localiza- CMV intron A sequence. The consensus of these findings tion of plasmids.43,44 Other viral promoter DNA se- is that direct DNA plasmid injection driven by the CMV quences, such as oriP which the EBV virus uses for promoter/enhancer together with its intronic A se- nuclear targeting, could also be beneficial.45,46 However, quences, and flanked by an SV40 enhancer with its poly not all viral promoter sequences have this capacity: A sequences is the most potent vector construct. Figure 2 the CMV promoter and the RSV LTR, for example, depicts this construct, which can serve as the basis for appear unable to directly target the plasmids into the further improvement in vector construction. Further nucleus. improvements might include nuclear localization ele- Expression of transgenes can also be enhanced by co- ments to enhance transport of plasmids into the nucleus expression of transactivators with recognition sequences as well as regulatory elements for export, stabilization built into the vector. These include serum response and translation of transcripts as detailed below. elements within CMV and some skeletal muscle promo- ters, such as alpha-actin promoter. Li et al reported that co-expression of serum response factor (SRF) increased Vector construction for enhancing nuclear transport luciferase gene expression by five fold or more when the and expression transgene was under control of skeletal muscle alpha- The potency of transgene expression partly depends on actin promoter and CMV promoter.47 the efficiency with which plasmid DNA is transported into nuclei. Transport of macromolecules between Inducible promoter systems cytoplasm and nucleoplasm is a dynamic process, The advance in vector construction has so far been mediated mainly via the nuclear pore complex by at limited mainly to improvement in efficiency of gene least three different classes of transport receptor sys- expression. Ideally, vector systems should be able to

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 135 confer a regulated expression required for different for transduction of muscle than the same plasmid clinical applications. An example of such needs is the delivered as naked DNA. The reasons for this ineffec- expression of insulin for the treatment of diabetes, which tiveness in skeletal muscle in vivo are not clear. One requires expression to be coordinated with blood glucose obvious difference between muscle fibres and the cells in levels. Another example is the expression of erythro- tissues such as liver, kidney and skin is that every muscle poietin (EPO) for the correction of anaemia associated fibre is surrounded by a layer of mechanically strong with diseases such as renal failure.48 extracellular matrix (ECM), a basement membrane rich Several systems have been developed for achieving in glycosaminoglycans (GAGs). It has been suggested regulated transgene expression.49–51 These approaches that directly injected lipoplexes may bind to those use the principle of controlling transgene expression by negatively charged ECM components.63 This is sup- means of elements or transcription factors designed ported by the work of Ruponen et al64 who demonstrated to bind selectively to the promoter which drives that the transfection efficiency of several commonly used the transgene. Expression of transgenes can thus be cationic lipid formulations, such as DOTAP, DOTAP/ regulated by modulating expression of these transcrip- Chol,54 DOTAP/DOPE,65 DOSPER,66 DOGS,67 can be tion factors or altering their activity through drug blocked by anionic heparan sulphate. Furthermore, the administration. However, regulation of transgene retardation of lipids in ECM may cause immune and expression in vivo by such approaches is unreliable, inflammatory responses, which could damage the mainly due to the leakage of activity in what ought adjacent transduced fibres. to be inactive states, and low levels of control associated Cationic polymers have also been disappointing for with the complexity of the systems. It is beyond the muscle transfection.68,69 The commonly used cationic scope of this review to discuss this issue in detail but polymers, poly-(L)-ornithine (PLO), polyamidoamine it is well covered in reviews by Mansuy et al50 and and branched polyethylemine (PEI) and its analogues Ozawa et al.51 mediate highly efficient plasmid delivery in vitro into many types of cells, including myoblasts.70–72 In vivo, Delivery system however, these polymers did not improve transfection efficiency in skeletal muscle. The reason(s) for the failure It is the inefficiency of naked plasmid DNA that restricts in vivo may again be the interaction between the its clinical potential as a vehicle for gene delivery. This positively charged polymers and negatively charged arises from a number of limitations most notable of ECM components within skeletal muscle. which are: instability in the extra-cellular milieu, poor In contrast, non-ionic polymers such as poly N-vinyl entry into the cell, tendency to be targeted to intracellular pyrrolidone (PVP) and some co-polymers have been degradation pathways and lack of effective transport into shown in vivo to enhance transduction of muscle fibres the nucleus. The majority of publications on this topic significantly.73 Based on the use of PVP, Rolland and have been dedicated to means of overcoming these colleagues established the concept of the ‘protective, various problems. We will discuss these approaches interactive, non-condensing’ (PINC) delivery system, individually, since many of them tackle more than one of and reported up to 10-fold enhancement of transgene the hurdles. expression over naked plasmid alone.73 They have used this delivery system to obtain high levels of biologically Lipids, liposomes and polymers active human growth hormone (hGH) in the circulation A number of cationic substances, such as liposomes, lipids following intramuscular injection.20 and polymers, have been investigated for their capacity to More recently, Lemieux et al reported that an amphi- improve efficiency of gene delivery.52–54 Cationic lipids philic carrier, SP1017, composed of a mixture of the block and liposomes are the most widely used and several co-polymers (poloxymers), pluronics L61 and F127, have been shown to raise the efficiency of in vitro gene significantly augmented intramuscular expression of delivery in many cell types. The mechanism(s) by which several reporter genes.68 SP1017 enhanced peak gene lipids and liposomes improve gene transfer is not clearly expression by about 10-fold compared with injected understood, but several factors, such as the type of cell naked plasmid alone and also led to sustained higher and their proliferative status, appear to be involved. The levels of expression. We too have observed an improve- positively charged lipids and liposomes are thought to ment in transgene expression in muscle with naked improve transgene delivery mainly through binding to plasmid delivered in conjunction with such a polymer and condensing negatively charged DNA, forming a formulation. An average of more than 200 muscle fibres complex called lipoplex in which the DNA is protected can be transduced to express the GFP marker gene at the against extracellular degradation. Moreover, the posi- site of single injection (Figure 3), compared with less tively charged lipoplex binds to the negatively charged than 10 fibres in control injections of the plasmid alone. cell surface molecules facilitating endocytosis. Once in An important advantage of these polymers is that they the endosome, some lipids may destabilize the endoso- exhibit a significantly higher efficiency, with an optimal mal membrane and encourage the release of DNA into dose of the carrier some 500-fold lower in amount than the cytosol, thus avoiding the lysosomal degradation PVP. The block co-polymers used in SP1017 are listed in pathway.55–57 Such lipoplexes can effectively deliver the US Pharmacopoeia as ‘inactive excipients’ and are transgenes to myoblasts in culture and many different widely used for drug delivery.74,75 Toxicity studies have types of cells in vivo, including epithelial cells and shown that the amount of SP1017 used in gene delivery hepatocytes.58–62 However, the effectiveness of lipoplexes has a greater than 1000-fold safety margin, implying that on gene transfer in muscle in vivo has been disappoint- they could readily be used in clinical trials. The ing, both as reported by other groups61 and in our mechanism(s) by which poloxamers enhance gene laboratory. In fact, lipoplexed DNA is often less efficient transfer is not clear. However, SP1017 has been shown

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 136 sequences of amino acids of viral proteins, there is clearly room for exploitation of these peptides to boost delivery of naked plasmids. Indeed, whole inactive viral particles or viral proteins, particularly those derived from adenoviruses, have been shown to enhance transgene expression.72,76,77 Although the results are encouraging, the amount of viral protein required presents an impediment to therapeutic usage of this approach for in vivo transfection. However, it may have some applicability for ex vivo gene therapy. For example, Campeau et al used this principle for delivery of the full- length dystrophin gene (12 kb in size) into myoblasts, which were then transplanted into muscle.72 Intensive effort has been focused therefore on the design of short synthetic peptides that would mimic the useful functions of viral proteins, while avoiding the disadvantages associated with virus infection, namely, induction of inflammation, immunogenicity and toxicity.78 Most peptides designed and tested so far have either originated from earlier studies on drug delivery or have been modified from viral protein sequences. The im- portant features of these polypeptides are thought to be their ability to bind and/or condense DNA, or to destabilize cell membranes (see review by Plank79). Poly-L-lysine and its derivatives are the most widely reported polypeptides employed for gene delivery. Positively charged lysine polymers bind strongly to negatively charged DNA, compacting it and possibly protecting it from enzyme degradation. But their ability to improve transfection on their own has proved to be limited.80 On the other hand, when such positively charged polymers are used in combination with mem- brane-destabilizing reagents such as inactive viral particles,77 viral fusion peptides,81,82 or receptor ligands such as transferrin,80 transfection efficiency increased Figure 3 Cross-sections of tibialis anterior muscle of C57 B10 mice significantly. Such combinations have not been tested for transfected with plasmid DNA containing a marker gene expressing gene transfer in muscle. cytoplasmic GFP. The plasmid DNA was injected intramuscularly with In search of peptides with selective membrane poloxamers in SP1017 formulation (a) or with saline as a control (b) in a destabilisation properties to improve gene delivery, m total volume of 30 l at a single site. Mice were killed and muscles Subbarao et al designed a 30-residue amphipathic sectioned 7 days after injection. GFP expression was viewed by 83 fluorescence microscopy. The background red colour shows the outlines peptide. This peptide, GALA (WEAALAEALAEA- of muscle fibres, whereas the green signals indicate fibres expressing GFP. LAEHLAEALAEALEALAAGGSC), has repeat units of More than 150 muscle fibres express GFP in this field alone (a). In glutamic acid–alanine–leucine–alanine and possesses the contrast, less than 10 fibres express GFP in the control muscle (b) injected property of destabilizing membranes in a pH-dependent with the same plasmid at the same concentration but in the absence of the manner. Parente et al examined the membrane-destabi- polymers. Note the variation in signal intensity among GFP-positive fibres. lizing effect of GALA peptide, on the experimental model of lipid biolayer vesicles of egg phosphatidylcho- line in a lipid/peptide molar ratio of 500/1.84 At pH 5.0, to considerably increase plasmid DNA diffusion within leakage of contents was detected from 100% of the muscle tissue.68 vesicles, but was abrogated at pH 7.5. This is thought to result from pH-dependent conformational changes of the Proteins and polypeptides peptide, which binds to the bilayer at pH 5.0, but not at The high efficiency of virus infection is thought in neutral pH. The drawback of this peptide is that it has general to be mediated by mechanisms that overcome the same charge as DNA and consequently binding of each of the four hurdles described above. First, the viral the two is weak, resulting in limited effects on gene envelope stabilizes the viral genome in extracellular delivery. However, covalent attachment of the peptide to compartments. Second, internalization of viral particles polyamidoleamine, via a disulphfide linkage, raises the via endosomal vacuoles is facilitated by binding of transfection efficiency of the 1:1 complex by 2–3 orders of specific viral surface proteins to cognate cell surface magnitude.85 This augmentation of transfection effi- receptors. Third, by destabilizing endosomal mem- ciency is likely due to the combination of the DNA branes, viral particles are released into the cytosol thus condensing and the pH buffering effect in the endosomal escaping the lysosomal degradation pathway. Finally, the compartment by polyamidoleamine together with the viral genomes gain access to the nucleoplasm by active destabilizing effect of the peptide on cell membranes. targeting, again through specific viral proteins. Since Again the effect of such types of peptide on gene these viral functions are executed by specific domains or delivery in muscle is untested.

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 137 As a development of this idea, Wyman and colleagues agents may lead to identification of optimal conditions designed a polypeptide called KALA (WEAKLAKALA- for gene delivery into the nuclei of specific cell types. KALAKHLAKALAKALKACEA), which differs from GALA principally in the replacement of the seven Electroporation uncharged A residues by positively charged K resi- Electropermeabilization, commonly referred to as elec- dues.86 KALA too is an amphipathic peptide: when in troporation, is a physical process which exposes cells to a alpha-helical conformation, one face displays hydropho- brief, high-intensity electric field that induces temporary bic leucine residues, while the opposite face displays damage to the plasma membrane, allowing the influx of hydrophilic lysine residues. The KALA peptide was large molecules to which it is normally impermeable. reported to cause membrane disruption as well as This technique has been used for nearly two decades for binding and condensing DNA. Transfection efficiency transfection of cells in vitro.95 Its initial application in vivo of non-myogenic cells with KALA/DNA complexes in was to transfer DNA into skin and liver cells, and was vitro is over 100 times greater than that with the optimal first reported for gene transfer into skeletal muscle in vivo poly-L-lysine/DNA complex delivery system. Relatively by Aihara and Miyazaki in 1998.96 Since then, progres- high transfection efficiency was also achieved in C2C12 sive improvements have made it the most efficient myoblasts and myotubes. However, the effect of KALA approach among non-viral gene delivery systems in on gene transfer in vivo in muscle has yet to be skeletal muscle. Applications of this technique have explored.87 been extended from treating muscular dystrophy to So far we have discussed enhancement of plasmid using muscle for systemic delivery of therapeutic entry into cells and their release from endosomes into the proteins.97–101 The mechanisms by which this method cytoplasm. Another objective in the development of non- enhances transgene expression appear to be relatively viral gene delivery systems is to improve the efficiency of simple. Electroporation is effective only when plasmid transport of plasmid DNA from the cytoplasm into DNA is injected into the muscle prior to, not after, the nucleus. This may be particularly important in electroporation. This suggests that the electric pulse acts muscle fibres, where, although plasmid entry may be directly on the charged DNA molecules and forces them relatively easier due to its vast surface area of T-tubule to migrate through ECM and cell membranes and enter system, translocation of naked plasmid to the nucleus muscle fibres, although membrane permeabilization may is likely to be more problematic. This is mainly also play a role.97 The entry of plasmid directly into because of the non-dividing nature of the myonuclei cytoplasm may bypass the endosome–lysosome path- which excludes the possibility of plasmid entry during way, reducing the degree of DNA degradation (Figure 4). mitosis. Thus plasmid DNA can gain access to myonuclei The effectiveness of gene transfer in muscle relies on only through the nuclear pore complex. This route of the widest possible dissemination of DNA or virus entry is standardly used by nuclear proteins such as within the area to be electrically pulsed. However, the transcription factors which are actively imported by abundant ECM in the perimysium and endomyosium of mediation of NLS within the cargo proteins. It is muscle restricts severely the diffusion of injected conceivable that a combination of NLS with plasmid plasmid, thus limiting the region over which fibres can DNA might improve efficiency of gene delivery (see be transduced. The pore size of basal lamina has also reviews78,88,89). Attempts have been made to either add been shown to selectively limit the penetration of larger nucleoproteins or NLS into cocktails of plasmid solution viruses, such as adenovirus, by comparison with the or to covalently link NLS to plasmid DNA, either smaller ones, such as AAV.102 Such size constraints may directly90 or indirectly through DNA-binding agents also be limiting factors for efficient delivery of large such as cationic polymers and peptides.91 Collas and plasmids. To improve dissemination of both viral and Alestrom in experiments involving micro-injection of non-viral vectors, some proteases have been tested for plasmid into the cytoplasm of zebrafish eggs, reported their ability to break down the barriers. To facilitate the that use of NLS increased nuclear accumulation of diffusion of AAV in muscles, Favre et al used hyalur- plasmid DNA and expression of the transgene.92 onidase which hydrolyses hyaluronic acid, a major Beneficial effects were also obtained in liver gene constituent of the muscle ECM, reporting up to a three delivery experiments in vivo93 and in different cell lines94 fold increase in infection efficiency.103 Likewise, Wells when NLS was used in combination with a liposome- and colleagues recently reported that pretreatment of mediated delivery system. However, little or no benefit muscle with hyaluronidase improves the penetration of over plasmid alone was reported when NLS was used in naked plasmid within the tissue and expression. To their combination with adenovirus/PEI for gene transfer into surprise, hyaluronidase pretreatment also reduced mus- myoblasts.72 cle damage.22 The limited benefit of simply adding NLS to a plasmid Although electroporation has shown high efficiency of DNA delivery system suggests the need for a more stable gene transfer in primates as well as in rodents, several linkage of NLS to plasmid DNA either directly or aspects need to be improved and the efficacy in human is indirectly through DNA-condensing and membrane- yet to be tested. The voltage gradient required to achieve destabilizing polymers. Again the success of this type efficient gene transfer, about 200 V/cm, has detrimental of approach has, so far, been limited (see review by effects on muscle fibres, provoking degeneration parti- Bremener et al78). This ineffectiveness may be attributable cularly in the centre of electric field. One can envisage to modifications of DNA and increase in size of plasmid/ the practical difficulties of adapting such a technique to polymer complexes, hampering the entry of DNA into the larger scales encountered in the human subject, cells. Clearly this approach to gene delivery is still in the particularly for gene therapy of genetic diseases, such as early stages of development and exhaustive testing of the the muscular dystrophies, which require body-wide variety of NLS sequences in combination with different delivery of therapeutic plasmids. Extensive tissue da-

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 138

Figure 4 Schematic diagram of the proposed routes by which naked plasmid DNA is delivered to muscle fibres and transgene products are secreted into circulation. The route of endocytosis (a) is exemplified by lipoplex-mediated plasmid delivery whereas the route of passive entry (b) is by electroporation. The efficiency of plasmid delivery into myonuclei through the route of endocytosis can be improved by modulation of plasmid, membrane targeting, endosome destabilization and nuclear targeting. Mechanisms which regulate transcription, translation of transgene and secretion of transgene products are largely unknown and have not been explored as a means of improvement in transgene expression.

mage may also act as a ‘Danger Signal’, promoting the current use of this technology is in targeting tumours. development of immune responses to transgene pro- Plasmid DNA or microspheres (such as microbubbles) ducts.104 Efficiency of gene transfer with electroporation bearing plasmids (MSP) can be injected intravenously also appears to be greatly dependent on the size of the and US energy applied to the target region. Secondly, the plasmid DNA; a 5 kb plasmid expressing b-gal can sonication needed to achieve enhanced gene expression transduce more than half the fibres (about 2000) of the by either plasmid alone or coupled to MSP is within the tibialis anterior muscle in a mouse while far fewer were range emitted by diagnostic transducers.107–109 Thus, if transduced with a 12.5 kb vector expressing human efficient gene transfer could be achieved by this dystrophin.105 This is clearly a major impediment to technique, it would have great advantages in clinical application of this technique to Duchenne’s muscular use over other physical methods, such as electroporation. dystrophy. Thirdly, enhanced transgene expression can be achieved by US alone or in combination with other non-viral Other approaches delivery systems, such as lipoplex-mediated gene deliv- Several other developments have emerged recently, ery. Lipoplexes in combination with US can enhance which have the potential to improve the efficiency of transgene expression by several orders of magni- naked plasmid DNA delivery systems both locally and tude.110,111 Manome et al reported that US of 20 W/cm2, systemically in muscles, and are worthy therefore of at continuous 1 MHz, increased b-galactosidase expres- discussion in brief. sion by 3–270-fold, in vitro and in vivo.112 In the presence of microbubbles, US enhanced gene expression up to Microbubbles and ultrasound 300-fold over plasmid alone. The main mechanism by Transfection of mammalian cells with plasmid DNA by which US enhances gene transfer is thought to be sonication was reported by Fechheimer et al as early as acoustic cavitation, which can effect transient non-lethal 1987.106 This technique has some advantages for gene perforations in the plasmalemma and possibly the delivery. Firstly, the energy of ultrasound (US) can be nuclear membranes.107 Microbubbles, by acting as focused on a relatively small area wherever gene cavitation nuclei, may potentiate pore formation in cell transduction is required. This is especially important membranes, thus facilitating the entry of plasmids for gene delivery into deeply located tissues. Hence the into cells and their release from endosomes.113 Our

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 139 preliminary results using US and microbubbles for gene the cytoplasmic accumulation of transgene mRNA.121 delivery in muscle also showed a significant increase in Zufferey et al reported a substantial increase in levels of the number of fibres transduced with a GFP marker gene expression in a transgene-, promoter- and vector-inde- (unpublished observation). pendent manner when a PRE of woodchuck hepatitis virus (WPRE) was inserted in the 30 untranslated region of coding sequences carried by either oncoretroviral or Local ischemia, vasodilation and high-pressure lentiviral vectors.122 injection Taking a different approach to this problem in AdV- As discussed earlier, one common hurdle for gene mediated gene transfer, it has been shown that tran- delivery in muscle, by either naked plasmid, electro- scriptionally active drugs such as retinoic acid (RA) in poration, microbubbles or virus infection, is the disper- combination with histone deacetylase inhibitor trichos- sion of the vehicles carrying transgenes within muscles. tatin A (TSA) can enhance and prolong transgene Several groups have tried to overcome this problem by expression in cell lines and skeletal muscle up to intravenous administration of the vehicles with local seven-fold over controls.123 This effect may be mediated, ischemia, vasodilation and high-pressure, or using high- in part, by the direct activation of RA receptors on the pressure injection. Takeshita et al reported that gene CMV promoter which contains repeated RA responsive expression in skeletal muscle transfection is significantly element (RARA). Perhaps more importantly, these drugs augmented by administration of transgenes under may reactivate expression of the transgene via remodel- ischemic conditions created by ligation of femoral artery ling of chromatin. This enhancing effect is unlikely to be of rat limb.114 Similarly, Wolff’s group reported wide- limited to the genes delivered by AdV. Indeed, expres- spread transgene expression when naked plasmid in a sion of a transgene, in vitro, delivered by polyoma virus large volume was delivered by rapid intra-arterial pseudocapsids can also be enhanced by these drugs (N injection in combination with occlusion of large draining Krauzewicz, personal communicaton, MRC clinical blood vessels.115,116 They also found moderate increases Science Centre, London, UK). We too have observed an in transgene expression when collagenase, papaverine increase in the number of myoblasts expressing a GFP and ischemia were used. These results suggest that local transgene when TSA and RA are added after PEI- ischemia, and more particularly the high pressure mediated gene transfer with naked plasmid DNA induced by vessel obstruction and high volume of (unpublished observation). The inclusion of PRE in transfection reagent, probably create enough force to vector constructs and use of drugs that activate tran- break ECM and membrane barriers and facilitate entry of scription might represent an important ancillary strategy plasmid into target cells. Similar protocols have also been for vector design and efficient gene expression. reported to be effective for gene delivery into the diaphragm117 and for virus-mediated gene delivery. Conclusion Greelish et al reported efficient delivery of human delta-sarcoglycan to distal hindlimb muscles through The efficiency of transgene delivery and expression by infusion of recombinant AAV into the femoral artery in non-viral delivery systems in muscle has been improved conjunction with histamine-induced endothelial permea- significantly. This is mainly due to the construction of bilization.118 new vectors, the use of gene delivery enhancing reagents Particle bombardment is well known to be effective for and the development of novel delivery systems. As immunization by gene therapy and is an alternative illustrated in Figure 4, nearly all of the measures taken so means of achieving forceful delivery of transgenes into far focus on tackling obstacles which prevent the efficient muscle. Stable and long-term expression of a marker entry of plasmid into cytosols and nuclei, or which cause gene in fish muscle was reported with naked plasmid DNA degradation. However, although it is known that delivered by gene gun.119 Walther et al used a low- plasmid DNA can persist in myonuclei for long periods, volume, high-speed hand-held jet injector for direct gene little is known about how the expression of transgenes in transfer of naked plasmid into tumours, and reported episomal DNA is controlled by cellular transcriptional that expression of the transgene was more efficient and and translational machineries. Nor do we know how the more widespread within tumour tissues than the control protein products of transgenes are transported from procedure of direct injection from a syringe.120 The muscle fibre cytoplasm into the circulation. The ultimate advantages of forceful delivery of transgenes in muscle efficiency of transgene expression and delivery of the with these methods remain to be widely tested. gene products could therefore be interfered with or improved at each of a number of stages. For example, Enhancing transgene expression at recent evidence suggests that active RNA polymerases transcription or post-transcription levels are concentrated in discrete ‘factories’ where they work together on many different templates. The evidence that Entry of transgenes into the nucleus does not guarantee such factories specialize in the transcription of particular efficient expression. It is well documented that plasmid groups of genes prompts the speculation that these DNA may persist within target cells with no detectable factories could be targeted for efficient transcription.124 expression.33 It is not well understood how transgene Although gene therapies have not lived up to hopes expression is regulated in muscle. Studies of transgene voiced for them 19 years ago,125 research in this topic has expression with viral infection systems suggest that there nonetheless made a number of useful advances. Princi- is room for improvement of transgene expression at both pal among these is the accurate definition of the the transcriptional and post-transcriptional levels. For biological hurdles that limit what is achievable and the instance, the post-transcriptional regulatory element identification of what appear to be realistic targets for (PRE) present in hepatitis B virus (HBV) can stimulate our current technologies. In this regard, skeletal muscle

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 140 has been shown repeatedly to be a robust target organ. 19 McMahon JM et al. Inflammatory responses following direct Gradual refinement of our understanding of how muscle injection of plasmid DNA into skeletal muscle. Gene Ther 1998; 5: fibres take up plasmid DNA and of the mechanisms that 1283–1290. mediate transcriptional and post-transcriptional controls 20 Fewell JG et al. Gene therapy for the treatment of hemophilia B of transgene expression look likely to lead to further using PINC-formulated plasmid delivered to muscle with improvements in efficiency of transgene expression to electroporation. Mol Ther 2001; 3: 574–583. the point where naked plasmid delivery would be a 21 Herzog RW et al. Long-term correction of canine hemophilia B genuine option for therapeutic application. This ap- by gene transfer of blood coagulation factor IX mediated by proach, avoiding viral vectors, has the virtue of adeno-associated viral vector. Nat Med 1999; 5: 56–63. minimizing contentiousness in the passage from animal 22 McMahon JM et al. Optimisation of electrotransfer of plasmid experiments to human trials. into skeletal muscle by pretreatment with hyaluronidase – increased expression with reduced muscle damage. Gene Ther 2001; 8: 1264–1270. References 23 Blaveri K et al. Patterns of repair of dystrophic mouse muscle: studies on isolated fibers. Dev Dyn. 1999; 216: 244–256. 1 Wolff JA et al. Direct gene transfer into mouse skeletal muscle in 24 Partridge TA. Invited review: myoblast transfer: a possible vivo. Science 1990; 247: 1465–1468. therapy for inherited myopathies? Muscle Nerve 1991; 14: 197– 2 Naldini L. Lentiviruses as gene transfer agents for delivery to 212. non-dividing cells. Curr Opin Biotechnol 1998; 9: 457–463. 25 Gussoni E et al. Dystrophin expression in the mdx mouse 3 Wang Y et al. HSV-1 amplicon vectors are a highly efficient gene restored by stem cell transplantation. Nature 1999; 401: 390–394. delivery system for skeletal muscle myoblasts and myotubes. 26 McKinney-Freeman SL et al. Muscle-derived hematopoietic stem Am J Physiol Cell Physiol 2000; 278: C619–C626. cells are hematopoietic in origin. Proc Natl Acad Sci USA 2002; 99: 4 Akkaraju GR et al. Herpes simplex virus vector-mediated 1341–1346. dystrophin gene transfer and expression in MDX mouse skeletal 27 Arruda VR et al. Protein posttranslational modifications of muscle. J Gene Med 1999; 1: 280–289. recombinant myotube-synthesized human factor IX. Blood 2001; 5 Tsukamoto H et al. Enhanced expression of recombinant 97: 130–138. dystrophin following intramuscular injection of Epstein–Barr 28 Tripathy SK et al. Long-term expression of erythropoietin in the virus (EBV)-based mini-chromosome vectors in mdx mice. Gene systemic circulation of mice after intramuscular injection of a Ther 1999; 6: 1331–1335. plasmid DNA vector. Proc Natl Acad Sci USA 1996; 93: 10 876– 6 Chao H et al. Several log increase in therapeutic transgene 10 880. delivery by distinct adeno-associated viral serotype vectors. Mol 29 Wells DJ, Goldspink G. Age and sex influence expression of Ther 2000; 102: 619–623. plasmid DNA directly injected into mouse skeletal muscle. FEBS 7 Xiao X, Li J, Samulski RJ. Efficient long-term gene transfer into Lett 1992; 306: 203–205. muscle tissue of immunocompetent mice by adeno-associated 30 Li X, Eastman EM, Schwartz RJ, Draghia-Akli R. Synthetic virus vector. J Virol 1996; 70: 8098–8108. muscle promoters: activities exceeding naturally occurring 8 Kessler PD et al. Gene delivery to skeletal muscle results in regulatory sequences. Nat Biotechnol 1999; 17: 241–245. sustained expression and systemic delivery of a therapeutic 31 Xu ZL et al. Optimization of transcriptional regulatory elements protein. Proc Natl Acad Sci USA 1996; 93: 14 082–14 087. for constructing plasmid vectors. Gene 2001; 272: 149–156. 9 Hagstrom JN et al. Improved muscle-derived expression of 32 Reecy JM et al. Multiple regions of the porcine alpha-skeletal human coagulation factor IX from a skeletal actin/CMV hybrid actin gene modulate muscle-specific expression in cell culture enhancer/promoter. Blood 2000; 95: 2536–2542. and directly injected skeletal muscle. Anim Biotechnol 1998; 9: 10 Wang B, Li J, Xiao X. Adeno-associated virus vector carrying 101–120. human minidystrophin genes effectively ameliorates muscular 33 Kelly R et al. Myosin light chain 3F regulatory sequences conter dystrophy in mdx mouse model. Proc Natl Acad Sci USA 2000; regionalized cardiac and skeletal muscle expression in trans- 97: 13 714–13 719. genic mice. J Cell Biol 1995; 129: 383–396. 11 Xiao X et al. Full functional rescue of a complete muscle (TA) in 34 Bou-Gharios G et al. Differential expression and secretion of dystrophic hamsters by adeno-associated virus vector-directed alpha 1 anti-trypsin between direct DNA injection and im- gene therapy. J Virol 2000; 74: 1436–1442. plantation of transfected myoblast. Gene Ther 1999; 6: 1021–1029. 12 Lu DR et al. Stage I clinical trial of gene therapy for hemophilia 35 Franz WM, Rothmann T, Frey N, Katus HA. Analysis of tissue- B. Sci China B 1993; 36: 1342–1351. specific gene delivery by recombinant adenoviruses containing 13 Kay MA et al. Evidence for gene transfer and expression of factor cardiac-specific promoters. Cardiovasc Res 1997; 35: 560–566. IX in haemophilia B patients treated with an AAV vector. Nat 36 Cordier L et al. Muscle-specific promoters may be necessary for Genet 2000; 24: 257–261. adeno-associated virus-mediated gene transfer in the treatment 14 Fields PA et al. Risk and prevention of anti-factor IX formation in of muscular dystrophies. Hum Gene Ther 2001; 12: 205–215. AAV-mediated gene transfer in the context of a large deletion of 37 Wang B, Li J, Xiao X. Adeno-associated virus vector carrying F9. Mol Ther 2001; 4: 201–210. human minidystrophin genes effectively ameliorates muscular 15 Drittanti L et al. Optimised helper virus-free production of high- dystrophy in mdx mouse model. Proc Natl Acad Sci USA 2000; quality adeno-associated virus vectors. J Gene Med 2001; 3: 59–71. 97: 13 714–13 719. 16 Nakai H, Storm TA, Kay MA. Increasing the size of rAAV- 38 Lu QL et al. Massive idiosyncratic exon skipping corrects the mediated expression cassettes in vivo by intermolecular joining nonsense mutation in dystrophic mouse muscle and produces of two complementary vectors. Nat Biotechnol 2000; 18: 527–532. functional revertant fibers by clonal expansion. J Cell Biol 2000; 17 Duan D, Yue Y, Yan Z, Engelhardt JF. A new dual-vector 148: 985–996. approach to enhance recombinant adeno-associated virus- 39 Barnhart KM et al. Enhancer and promoter chimeras in plasmids mediated gene expression through intermolecular cis activation. designed for intramuscular injection: a comparative in vivo and Nat Med 2000; 6: 595–598. in vitro study. Hum Gene Ther 1998; 9: 2545–2553. 18 Sun L, Li J, Xiao X. Overcoming adeno-associated virus vector 40 Li S et al. Muscle-specific enhancement of gene expression by size limitation through viral DNA heterodimerization. Nat Med incorporation of SV40 enhancer in the expression plasmid. Gene 2000; 6: 599–602. Ther 2001; 8: 494–497.

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 141 41 Ribbeck K, Gorlich D. Kinetic analysis of translocation through 64 Ruponen M et al. Extracellular glycosaminoglycans modify nuclear pore complexes. EMBO J 2001; 20: 1320–1330. cellular trafficking of lipoplexes and polyplexes. J Biol Chem 42 Chan CK, Hubner S, Hu W, Jans DA. Mutual exclusivity of DNA 2001; 276: 33 875–33 880. binding and nuclear localization signal recognition by the yeast 65 Leventis R, Silvius JR. Interactions of mammalian cells with lipid transcription factor GAL4: implications for nonviral DNA dispersions containing novel metabolizable cationic amphi- delivery. Gene Ther 1998; 5: 1204–1212. philes. Biochim Biophys Acta 1990; 1023: 124–132. 43 Dean DA. Import of plasmid DNA into the nucleus is sequence 66 Dodds E et al. Lipofection of cultured mouse muscle cells: a specific. Exp Cell Res 1997; 230: 293–302. direct comparison of Lipofectamine and DOSPER. Gene Ther 44 Dean DA, Dean BS, Muller S, Smith LC. Sequence require- 1998; 5: 542–551. ments for plasmid nuclear import. Exp Cell Res 1999; 253: 67 Behr JP. Gene transfer with synthetic cationic amphiphiles: 713–722. prospects for gene therapy. Bioconjug Chem 1994; 5: 382–389. 45 Langle-Rouault F et al. Up to 100-fold increase of apparent gene 68 Lemieux P et al. A combination of poloxamers increases gene expression in the presence of Epstein-Barr virus oriP sequences expression of plasmid DNA in skeletal muscle. Gene Ther 2000; 7: and EBNA1: implications of the nuclear import of plasmids. J 986–991. Virol 1998; 72: 6181–6185. 69 Mumper RJ et al. Polyvinyl derivatives as novel interactive 46 Tsujie M et al. Prolonged transgene expression in glomeruli polymers for controlled gene delivery to muscle. Pharm Res 1996; using an EBV replicon vector system combined with HVJ 13: 701–709. liposomes. Kidney Int 2001; 59: 1390–1396. 70 Nguyen HK et al. Evaluation of polyether–polyethyleneimine 47 Li S et al. Increased level and duration of expression in muscle by graft copolymers as gene transfer agents. Gene Ther 2000; 7: 126– co-expression of a transactivator using plasmid systems. Gene 138. Ther 1999; 6: 2005–1123. 71 Jeong JH et al. DNA transfection using linear poly(ethylenimine) 48 Bohl D, Naffakh N, Heard JM. Long-term control of prepared by controlled acid hydrolysis of poly(2-ethyl-2-oxazo- erythropoietin secretion by doxycycline in mice trans- line). J Control Release 2001; 73: 391–399. planted with engineered primary myoblasts. Nat Med 1997; 3: 72 Campeau P et al. Transfection of large plasmids in primary 299–305. human myoblasts. Gene Ther 2001; 8: 1387–1394. 49 Ye X et al. Regulated delivery of therapeutic proteins after in vivo 73 Anwer K et al. Synergistic effect of formulated plasmid and somatic cell gene transfer. Science 1999; 283: 88–91. needle-free injection for genetic vaccines. Pharm Res 1999; 16: 50 Mansuy IM, Bujard H. Tetracycline-regulated gene expression in 889–895. the brain. Curr Opin Neurobiol 2000; 10: 593–596. 74 Batrakova EV et al. Pluronic P85 enhances the delivery of digoxin 51 Ozawa CR, Springer ML, Blau HM. A novel means of to the brain: in vitro and in vivo studies. J Pharmacol Exp Ther drug delivery: myoblast-mediated gene therapy and regu- 2001; 296: 551–557. latable retroviral vectors. Annu Rev Pharmacol Toxicol 2000; 40: 75 von Moltke LL, Greenblatt DJ. Drug transporters revisited. J Clin 295–317. Psychopharmacol 2001; 21: 1–3. 52 Cao X et al. Efficient inducation of local and systemic antitumor 76 Krauzewicz N et al. Sustained ex vivo and in vivo transfer of a immune response by liposome-mediated intratumoral co-trans- reporter gene using polyoma virus pseudocapsids. Gene Ther fer of interleukin-2 gene and interleukin-6 gene. J Exp Clin Cancer 2000; 7: 1094–1102. Res 1999; 18: 191–200. 77 Kremer EJ, Perricaudet M. Adenovirus and adeno-associated 53 Vitiello L, Bockhold K, Joshi PB, Worton RG. Transfection of virus mediated gene transfer. Br Med Bull 1995; 51: 31–44. cultured myoblasts in high serum concentration with DODAC:- 78 Bremner KH, Seymour LW, Pouton CW. Harnessing nuclear DOPE liposomes. Gene Ther 1998; 5: 1306–1313. localization pathways for transgene delivery. Curr Opin Mol Ther 54 Felgner PL et al. Lipofection: a highly efficient, lipid-mediated 2001; 3: 170–177. DNA-transfection procedure. Proc Natl Acad Sci USA 1987; 84: 79 Plank C, Zauner W, Wagner E. Application of membrane-active 7413–7417. peptides for drug and gene delivery across cellular membranes. 55 Stamatatos L, Leventis R, Zuckermann MJ, Silvius JR. Interac- Adv Drug Deliv Rev 1998; 34: 21–35. tions of cationic lipid vesicles with negatively charged phos- 80 Wagner E et al. DNA-binding transferrin conjugates as func- pholipid vesicles and biological membranes. Biochemistry 1988; tional gene-delivery agents: synthesis by linkage of polylysine or 27: 3917–3925. ethidium homodimer to the transferrin carbohydrate moiety. 56 Zabner J et al. Cellular and molecular barriers to gene transfer by Bioconjug Chem 1991; 2: 226–231. a cationic lipid. J Biol Chem 1995; 270: 18 997–19 007. 81 Plank C et al. The influence of endosome-disruptive peptides on 57 Hafez IM, Maurer N, Cullis PR. On the mechanism whereby gene transfer using synthetic virus-like gene transfer systems. J cationic lipids promote intracellular delivery of polynucleic Biol Chem 1994; 269: 12 918–12 924. acids. Gene Ther 2001; 8: 1188–1196. 82 Gottschalk S et al. A novel DNA-peptide complex for efficient 58 Hyde SC et al. Repeat administration of DNA/liposomes to the gene transfer and expression in mammalian cells. Gene Ther nasal epithelium of patients with cystic fibrosis. Gene Ther 2000; 1996; 3: 48–57. 7: 1156–1165. 83 Subbarao NK et al. pH-dependent bilayer destabilization by an 59 Boussif O et al. Enhanced in vitro and in vivo cationic lipid- amphipathic peptide. Biochemistry 1987; 26: 2964–2972. mediated gene delivery with a fluorinated glycerophosphoetha- 84 Parente RA, Nadasdi L, Subbarao NK, Szoka Jr FC. Association nolamine helper lipid. J Gene Med 2001; 3: 109–114. of a pH-sensitive peptide with membrane vesicles: role of amino 60 Koehler DR et al. Targeting transgene expression for cystic acid sequence. Biochemistry 1990; 29: 8713–8719. fibrosis gene therapy. Mol Ther 2001; 4: 58–65. 85 Haensler J, Szoka Jr FC. Polyamidoamine cascade polymers 61 Helbling-Leclerc A, Scherman D, Wils P. Cellular uptake of mediate efficient transfection of cells in culture. Bioconjug Chem cationic lipid/DNA complexes by cultured myoblasts and 1993; 4: 372–379. myotubes. Biochim Biophys Acta 1999; 1418: 165–175. 86 Wyman TB et al. Design, synthesis, and characterization of a 62 Manthorpe M et al. Gene therapy by intramuscular injection of cationic peptide that binds to nucleic acids and permeabilizes plasmid DNA: studies on firefly luciferase gene expression in bilayers. Biochemistry 1997; 36: 3008–3017. mice. Hum Gene Ther 1993; 4: 419–431. 87 Rittner K et al. New basic membrane-destabilizing peptides for 63 Caron NJ et al. Intracellular delivery of a Tat-eGFP fusion protein plasmid-based gene delivery in vitro and in vivo. Mol Ther 2002; into muscle cells. Mol Ther 2001; 3: 310–318. 5: 104–114.

Gene Therapy Non-viral gene delivery in skeletal muscle QL Lu et al 142 88 Morris MC, Chaloin L, Heitz F, Divita G. Translocating peptides 107 Lawrie A et al. Microbubble-enhanced ultrasound for vascular and proteins and their use for gene delivery. Curr Opin Biotechnol gene delivery. Gene Ther 2000; 7: 2023–2027. 2000; 11: 461–466. 108 Newman CM, Lawrie A, Brisken AF, Cumberland DC. 89 Luo D, Saltzman WM. Synthetic DNA delivery systems. Nat Ultrasound gene therapy: on the road from concept to reality. Biotechnol 2000; 18: 33–37. Echocardiography 2001; 18: 339–347. 90 Ciolina C et al. Coupling of nuclear localization signals to 109 Amabile PG et al. High-efficiency endovascular gene delivery via plasmid DNA and specific interaction of the conjugates with therapeutic ultrasound. J Am Coll Cardiol 2001; 37: 1975–1980. importin alpha. Bioconjug Chem 1999; 10: 49–55. 110 Koch S, Pohl P, Cobet U, Rainov NG. Ultrasound enhancement 91 Godbey WT, Wu KK, Hirasaki GJ, Mikos AG. Improved packing of liposome-mediated cell transfection is caused by cavitation of poly(ethylenimine)/DNA complexes increases transfection effects. Ultrasound Med Biol 2000; 26: 897–903. efficiency. Gene Ther 1999; 6: 1380–1388. 111 Anwer K et al. Ultrasound enhancement of cationic lipid- 92 Collas P, Alestrom P. Rapid targeting of plasmid DNA to mediated gene transfer to primary tumors following systemic zebrafish embryo nuclei by the nuclear localization signal of administration. Gene Ther 2000; 7: 1833–1839. SV40 T antigen. Mol Mar Biol Biotechnol 1997; 6: 48–58. 112 Manome Y, Nakamura M, Ohno T, Furuhata H. Ultrasound 93 Kaneda Y, Iwai K, Uchida T. Increased expression of DNA facilitates transduction of naked plasmid DNA into colon cointroduced with nuclear protein in adult rat liver. Science 1989; carcinoma cells in vitro and in vivo. Hum Gene Ther 2000; 11: 243: 375–378. 1521–1528. 94 Fritz JD, Herweijer H, Zhang G, Wolff JA. Gene transfer into 113 Wei K et al. Interactions between microbubbles and ultrasound: mammalian cells using histone-condensed plasmid DNA. Hum in vitro and in vivo observations. J Am Coll Cardiol 1997; 29: Gene Ther 1996; 7: 1395–1404. 1081–1088. 95 Wong TK, Neumann E. Electric field mediated gene transfer. 114 Takeshita S, Isshiki T, Sato T. Increased expression of direct Biochem Biophys Res Commun 1982; 107: 584–587. gene transfer into skeletal muscles observed after acute 96 Aihara H, Miyazaki J. Gene transfer into muscle by electropora- ischemic injury in rats. Lab Invest 1996; 74: 1061–1065. tion in vivo. Nat Biotechnol 1998; 16: 867–870. 115 Zhang G et al. Efficient expression of naked dna delivered 97 Mir LM et al. High-efficiency gene transfer into skeletal muscle intraarterially to limb muscles of nonhuman primates. Hum mediated by electric pulses. Proc Natl Acad Sci USA 1999; 96: Gene Ther 2001; 12: 427–438. 4262–4267. 116 Budker V et al. The efficient expression of intravascularly 98 Yin D, Tang JG. Gene therapy for streptozotocin-induced delivered DNA in rat muscle. Gene Ther 1998; 5: 272–276. diabetic mice by electroporational transfer of naked human 117 Liu F, Nishikawa M, Clemens PR, Huang L. Transfer of full- insulin precursor DNA into skeletal muscle in vivo. FEBS Lett length Dmd to the diaphragm muscle of Dmd(mdx/mdx) mice 2001; 495: 16–20. through systemic administration of plasmid DNA. Mol Ther 99 Watanabe K et al. Protection against autoimmune myocarditis by 2001; 4: 45–51. gene transfer of interleukin-10 by electroporation. Circulation 118 Greelish JP et al. Stable restoration of the sarcoglycan complex 2001; 104: 1098–1100. in dystrophic muscle perfused with histamine and a recombi- 100 Maruyama H et al. Long-term production of erythropoietin nant adeno-associated viral vector. Nat Med 1999; 5: 439–443. after electroporation-mediated transfer of plasmid DNA into 119 Lee JY, Hirono II, Aoki T. Stable expression of a foreign gene, the muscles of normal and uremic rats. Gene Ther 2001; 8: 461– delivered by gene gun, in the muscle of rainbow trout 468. Oncorhynchus mykiss. Mar Biotechnol 2000; 2: 254–258. 101 Kishida T et al. In vivo electroporation-mediated transfer of 120 Walther W et al. Nonviral in vivo gene delivery into tumors interleukin-12 and interleukin-18 genes induces significant using a novel low volume jet-injection technology. Gene Ther antitumor effects against melanoma in mice. Gene Ther 2001; 2001; 8: 173–180. 8: 1234–1240. 121 Huang ZM, Yen TS. Role of the hepatitis B virus posttranscrip- 102 Pruchnic R et al. The use of adeno-associated virus to tional regulatory element in export of intronless transcripts. circumvent the maturation-dependent viral transduction of Mol Cell Biol 1995; 15: 3864–3869. muscle fibers. Hum Gene Ther 2000; 11: 521–536. 122 Zufferey R, Donello JE, Trono D, Hope TJ. Woodchuck hepatitis 103 Favre D et al. Hyaluronidase enhances recombinant adeno- virus posttranscriptional regulatory element enhances expres- associated virus (rAAV)- mediated gene transfer in the rat sion of transgenes delivered by retroviral vectors. J Virol 1999; skeletal muscle. Gene Ther 2000; 7: 1417–1420. 73: 2886–2892. 104 Matzinger P. Tolerance, danger, and the extended family. Annu 123 Gaetano C et al. Transcriptionally active drugs improve Rev Immunol 1994; 12: 991–1045. adenovirus vector performance in vitro and in vivo. Gene Ther 105 Vilquin JT et al. Electrotransfer of naked DNA in the skeletal 2000; 7: 1624–1630. muscles of animal models of muscular dystrophies. Gene Ther 124 Pombo A et al. Specialized transcription factories within 2001; 8: 1097–1107. mammalian nuclei. Crit Rev Eukaryot Gene Expr 2000; 10: 21–29. 106 Fechheimer M et al. Transfection of mammalian cells with 125 Friedmann T. Gene therapy – fact and fiction. In: Friedmann T plasmid DNA by scrape loading and sonication loading. Proc (ed.). Biology’s New Approaches to Diseases. Cold Spring Harbor Natl Acad Sci USA 1987; 84: 8463–8467. Laboratory, NY, 1983.

Gene Therapy