Enhanced expression of recombinant dystrophin following ... - Nature

Gene Therapy (1999) 6, 1331–1335  1999 Stockton Press All rights reserved 0969-7128/99 $12.00 http://www.stockton-press.co.uk/gt

BRIEF COMMUNICATION

Enhanced expression of recombinant dystrophin following intramuscular injection of Epstein–Barr virus (EBV)-based mini-chromosome vectors in mdx mice H Tsukamoto1,2, DJ Wells3, SC Brown1, P Serpente1, PN Strong4, J Drew1, K Inui2, S Okada2 and G Dickson1 1

Division of Biochemistry, School of Biological Sciences, Royal Holloway College, University of London, Egham, UK; 2Department of Pediatrics, Osaka University Medical School, Suita, Osaka, Japan; 3Gene Targeting Unit, Department of Neuromuscular Disease, Imperial College School of Medicine, London; and 4Division of Biomedical Sciences, Sheffield Hallam University, Sheffield, UK

Gene transfer by direct intramuscular injection of naked plasmid DNA has been shown to be a safe, simple but relatively inefficient method for gene delivery in vivo. Eukaryotic plasmid expression vectors incorporating the Epstein–Barr virus (EBV) origin of replication (oriP) and EBNA1 gene have been shown to act as autonomous episomally replicating gene transfer vectors which additionally provide nuclear matrix retention functions. Prolonged expression of a LacZ reporter gene and recombinant human dystrophin was shown using EBV-based plasmid vectors transfected into C2C12 mouse myoblast and

myotube cultures. Intramuscular injection of EBV-based dystrophin expression plasmids into nude/mdx mice resulted in significant enhancement in the number of muscle fibres expressing recombinant dystrophin compared with a conventional vector. This effect was observed for over 10 weeks after a single administration. These results indicate the potential advantage of EBV-based expression vectors for focal plasmid-mediated gene augmentation therapy in Duchenne muscular dystrophy (DMD) and a range of other gene therapeutic applications.

Keywords: gene therapy; muscle; dystrophin; EBNA-1; Epstein–Barr virus vector; Duchenne muscular dystrophy

Introduction Duchenne muscular dystrophy (DMD) is an X-linked recessive disorder affecting one in 3500 male births.1 The gene mutated in DMD and its milder form of Becker muscular dystrophy (BMD) encodes dystrophin, a large cytoskeletal protein of the spectrin superfamily. Mutations in the dystrophin gene result in the absence of dystrophin and dystrophin-associated proteins in skeletal muscle and other tissues. The early stages of DMD are characterised by repeated cycles of muscle fibre degeneration and regeneration. However, phagocytic infiltration and the replacement of myofibres by adipose and connective tissues are prominent features of the more advanced stages of the disease and result in a disruption of muscle architecture and function.1,2 Such is the course of the disease that DMD patients are usually wheelchair-bound before 12 years of age and die in their early twenties because of muscle weakness and atrophy. There is as yet no effective therapy for DMD, however, since this disease is caused by the mutation of a single recessive gene it may be considered as a candidate for gene therapy.3–6 Various studies have shown that full-length and Correspondence: JG Dickson, Division of Biochemistry, School of Biological Sciences, Royal Holloway-University of London, Egham Hill, Egham, Surrey, TW20 0EX, UK Received 31 December 1998; accepted 8 March 1999

Becker-type mini-dystrophin (6.3 kb) cDNA gene transfer is able to prevent the onset of dystrophy in dystrophindeficient mdx mice.3,7–10 The mdx mouse has a point mutation in the dystrophin gene which results in the virtual complete absence of the protein. Using adenoviral and retroviral vectors for gene delivery several studies have shown that both full-length and truncated forms of dystrophin are able to locate to the sarcolemma of muscle fibres of mdx mice.4,11–13 Viral vectors do, none the less, have several disadvantages with respect to safety and efficacy. In contrast, nonviral gene delivery systems offer the advantage of relative simplicity of preparation, avoidance of the size constraints of viral packaging, reduced toxicity and immunogenicity, and a generally advantageous safety profile. The main disadvantage of nonviral gene delivery, however, is the relative low efficiency of these methods in general. Development of delivery systems including direct plasmid DNA injection,7 lipofection,14,15 HVJ-liposomes,16 ligand-directed gene delivery and particle bombardment,17 have attempted to overcome this. In addition to problems of delivery, the plasmid-based eukaryotic expression vectors conventionally used do not replicate and so are not retained in the nucleus of rapidly dividing cells, with plasmid DNA being rapidly diluted.18,19 Furthermore, loss of plasmid DNA may also be unavoidable even in nondividing cells due to lack of nuclear matrix retention, and subsequent degradation and loss through the nuclear envelope.

EBV-based dystrophin vectors in mdx mice H Tsukamoto et al

1332

Epstein–Barr virus (EBV)-based episomal artificial chromosome vectors are nonviral gene transfer plasmid vectors carrying the EBNA-1 gene and the EBV origin for plasmid replication (oriP) as trans and cis elements for DNA replication.20 In the context of a eukaryotic expression plasmid, the oriP and EBNA1 elements confer the functions of autonomous episomal replication and nuclear retention in primate cells, but generally only the latter function in rodent cells.21 EBV-based vectors have thus been widely applied for gene transfer studies in human cells but much less so in rodent cells. However, in the presence of EBNA-1 protein, the repeat components of the EBV oriP are able to prolong the retention of replicating or nonreplicating vectors in dividing rodent cells.22 This nuclear retention function is thought to reflect an interaction between EBNA-1, oriP-containing

plasmid DNA, and the nuclear matrix chromosomal scaffold and appears to be sufficient to hold plasmid DNA in the nucleus.23 In primate cells EBV-based vectors replicate autonomously as an episome, and thereby offer the prospect of stable gene transfer without the necessity for integration into the host genome. The increase in the frequency of stable transfection observed with this type of vector is thus due to their maintenance as stable episomal plasmid DNA at two to 50 copies per cell.20 However, EBV-based plasmid vectors do incur some of the disadvantages of viral vectors with respect to safety. Viral origins of replication require at least one viral protein that may be toxic, oncogenic or immunogenic and often fail to operate correctly under situations of normal cell cycle control as in the cases of SV40 and polyoma-based vectors.21,24,25 Multiple replication cycles may give rise to deletions, recombinations or rearrangements of an unstable plasmid, and increased mutation frequency, as in the case of bovine papilloma virus (BPV)based replicating vectors.26 However, EBV appears to replicate only once per cell cycle, genome mutations are rare and EBNA-1 is nontoxic and non-immunogenic when expressed at low levels.21 Thus, episomal replicating vectors based on EBV have been widely proposed for employment in gene therapy protocols because of attractive toxicity and immunogenicity profiles coupled to dramatic improvement in transgene retention and longevity of expression. This article describes enhanced long-term expression of EBV-based vectors carrying dystrophin transgenes by mouse muscle cells following intramuscular injection of naked plasmid DNA. An initial study was undertaken to examine the activity of the CMV, RSV and PGK promoters in cultured mouse muscle cells in the context of the EBV-based and conventional LacZ reporter plasmids. In comparative studies the RSV promoter exhibited the highest activity of LacZ transgene expression in C2C12 mouse myotubes (not shown). To characterise and quantify the effects of inclusion of the EBV oriP/EBNA elements on expression of the plasmid transgenes in muscle and non-muscle cells, in vitro transfections with pRSV␤ and pRSV␤ + EB Figure 1 Prolongation of LacZ transgene expression in C2C12 myotubes following transfection with EBV-based expression plasmids. 293 Cells (A), NIH3T3 cells (B), and C2C12 cultures (C) were transfected with pRSV␤ (open bars) or pRSV␤+EB (filled bars), harvested after 16 h, 5 days and 9 days, and assayed for ␤-gal-specific activity. Values for each series at the 5 day and 9 day time-points are normalised and presented as % of those obtained at the 16 h time-point. Values are presented as means ± s.d. of independent cultures (n = 3). C2C12 myotube cultures were also stained for LacZ expression 9 days after transfection with pRSV␤ (D) or pRSV␤+EB (E). Plasmid pRSV␤ was constructed by removing the CMV promoter from pCMV␤ (Clontech Laboratories, Basingstoke, UK) using EcoRI and XhoI and then replacing it with the Rous sarcoma virus (RSV) promoter (SalI–XhoI fragment of pREP10; Invitrogen, Leek, The Netherlands). Plasmid pCMV␤+EB was constructed by blunt-end cloning a 4.2 kb XbaI–ClaI fragment from pREP10 containing the EBV oriP and EBNA-1 into a unique HindIII site downstream of the SV40 polyadenylation sites in pRSV␤. Plasmid transfection of cultures (5 × 105 cells seeded into 40 mm multiwell plates) was carried out using LipofectAMINE (Life Technologies, Paisley, UK).14,15 Cultures were maintained in DMEM with 10% FCS. In the case of C2C12 cells, medium was changed to DMEM plus 2% horse serum 1 day after transfection to promote formation of myotubes.14,15,37 Transfected NIH3T3 and 293 cells were sub-cultured (1:6 split) 1 day after transfection and then every subsequent 3 days. At various stages cultures were assayed for ␤-galactosidase (Stratagene, Amsterdam, The Netherlands) according to the manufacturer’s instructions, or stained for LacZ histochemistry.14


plasmids were performed in 293 (Figure 1A), NIH3T3 (Figure 1B) and C2C12 (Figure 1C, D and E)) cells. Cultures were then harvested and ␤-galactosidase specific activity determined at days 5 and 9 as a percentage of that obtained after 16 h. In the case of human 293 cells, EBV-based vectors should exhibit both episomal amplification and nuclear retention activities. Transfection with pRSV␤+EB resulted in expression levels of 98% and 27% at days 5 and 9, whereas with control pRSV␤ these values were 43% and 8% (Figure 1A). With mouse NIH3T3 cells, EBV-based vectors should exhibit nuclear retention activity but not episomal amplification. As shown in Figure 1B, ␤-galactosidase activity in transfected NIH3T3 cells at days 5 and 9 were 78% and 10% with pRSV␤+EB compared with 21% and 2% with the control plasmid pRSV␤. A contrasting situation was observed with the differentiating mouse C2C12 myoblast/myotube cultures (Figure 1C). At the day 5 time-point similar levels of ␤galactosidase activity were observed with both the pRSV␤ and pRSV␤+EB plasmids (42 and 44%, respectively). In the case of pRSV␤+EB transfected cells this level was maintained up to day 9, whereas with the control non-EBV plasmid, pRSV␤ the ␤-galactosidase lev-

els showed a significant drop. The increased level of LacZ transgene activity at day 9 in C2C12 myotube cultures transfected with pRSV␤+EB as compared with pRSV␤ was confirmed by histochemical staining (Figure 1D and E). Dystrophin expression plasmids utilising the RSV promoter with or without the EBV oriP/EBNA elements (pRSVDyB(N) and pRSVDyB(N)+EB) were then constructed based upon a 6.3 kb mini-dystrophin cDNA previously characterised in a variety of somatic and germline gene transfer studies in the mdx mouse.3,9,10 Dystrophin immunolabelled 293 cells and C2C12 myotubes were clearly observed following transfection with both the pRSVDyB(N) and pRSVDyB(N)+EB plasmids (not shown). In 293 cells, approximately twice as many positive cells were found with the oriP/EBNA-containing plasmid compared with the control vector. No qualitative or quantitative differences were discerned in the expression pattern of recombinant dystrophin in C2C12 myotubes transfected with pRSVDyB(N) or pRSVDyB(N)+EB plasmids, but in both cases localisation of immunolabelling to the plasma membrane was confirmed by confocal laser-scanning microscopy.

Figure 2 Enhanced recombinant dystrophin expression following direct intramuscular injection of EBV-based plasmid DNA in nude/mdx mice. Equimolar doses of mini-dystrophin expression plasmids pRSVDyB(N) (37 ␮g in 25 ␮l: A and open bars in C) or pRSVDyB(N)+EB (50 ␮g in 25 ␮l: B and filled bars in C) were injected into tibialis anterior muscles of 8- to 10-week-old nude/mdx mice.12,18,19 Plasmid DNAs were purified using Qiagen EndoFree plasmid kits (Qiagen, Crawley, UK), prepared in sterile PBS at 1–2 mg/ml and injected intramuscularly 5 days after BaCl2-induced regeneration.18,19 Transfected muscles were harvested after 1, 4 and 10 weeks and frozen sections processed to detect expression of recombinant dystrophin by immunolabelling with antibody, P6.10,38 Examples of immunolabelled sections are shown for the 4 week time-points (A and B: scale bar represents 100 ␮m). Fibres expressing dystrophin were counted over entire cross-sections of tibialis anterior muscles, and values at 1, 4 and 10 weeks are shown in C as means ± s.d. of groups of four treated animals.

1333


1334

To evaluate the use of EBV-based plasmid transfection of muscle in vivo in a relevant pre-clinical disease model, recombinant dystrophin expression in nude/mdx mice was examined following intramuscular injection of the pRSVDyB(N) and pRSVDyB(N)+EB plasmids. Equimolar doses of the EBV-based and control plasmids were injected into BaCl2 regenerated TA muscle in age- and sex-matched groups of animals, and treated muscle tissues examined by dystrophin immunocytochemistry after 1, 4 and 10 weeks using the anti-dystrophin polyclonal antiserum. Specific sarcolemmal immunolabelling corresponding to recombinant dystrophin was obtained in approximately 9–14% of fibres at all time-points with both plasmids (shown at 4 weeks in Figure 2A and B). The number of transfected fibres was consistently higher in the pRSVDyB(N)+EB-injected compared with the pRSVDyB-injected muscles at all three points (Figure 2C): statistical significance between the two plasmid treatments was P ⬍ 0.05 at the 1 and 4 week time-points and P ⬍ 0.1 at the 10 week time-point. The number of dystrophin-positive fibres was stably maintained across the 10 weeks of the experiment, and haematoxylin and eosin staining showed no atypical findings (data not shown). Skeletal muscle is an attractive gene therapy target not only for treatment of myopathies such as Duchenne muscular dystrophy but also as a platform tissue for ectopic expression of heterologous gene products destined to be secreted into the circulation. In both scenarios, high level and persistent transgene expression levels are crucial ingredients of any prospective therapeutic strategy. Here, we have evaluated in muscle gene transfer studies, the use of plasmid eukaryotic expression vectors harbouring the cis- and trans-acting elements of EBV which are responsible for replication and nuclear retention of episomal DNA. EBV-based vectors have already been used successfully in human cells for high efficiency in a number of gene transfection application.27–30 As far as we know, this is the first report of in vivo injection of EBVbased plasmid vectors into muscle in vivo. The EBV origin of replication is usually thought to function only in primate cells although there have been two reports describing the successful use of EBV-based vectors in rat glioblastoma cells31 and pheochromocytoma PC12 cells.32 EBNA-1 protein acts as a transcriptional regulator/ enhancer through binding to oriP, but the two elements in conjunction can mediate the physical retention of replicating or nonreplication vectors in dividing and nondividing cells via interaction with the nuclear matrix scaffolding.22 There have been several reports that EBV-based vectors carrying replicon sequences from genomic DNA provide efficient autonomous replication in a variety of mammalian cells,33,34 and that even plasmid vector lacking mammalian origins of replication exhibit EBNA-1mediated nuclear retention in both rodent and primate cells.34 In the present study, the nuclear retention functions of EBNA-1 and oriP appear to have been functionally active in mouse C2C12 and NIH3T3 cells in vitro, and in mouse muscle in vivo. Normal muscle tissue is a highly stable post-mitotic cellular system, but in the mdx mouse persistent degeneration/regeneration is ongoing up to a year of age.35,36 Under these conditions the nuclear retention functions of EBV-based vectors may be particularly advantageous for the maintenance of expression. As is the case in mdx mice, human DMD muscle exhibits exten-

sive muscle regeneration and hypertrophy except at the later stages of the disease. Thus this muscle may be an ideal target for the direct intramuscular injection of EBVbased dystrophin vectors. The ability of EBV-based vectors to replicate in muscle was not directly addressed in this study because no selection markers were used. Given that episomal replication is restricted to primate cells, it will be of considerable interest to test EBV-based vectors in human muscle cells in vitro. The potential therapeutic use of EBV-based dystrophin plasmid vectors by intramuscular injection at selected focal sites such as the diaphragm may be much more effective in human DMD patients than conventional plasmids. Recently, we have reported the successful direct injection of a mini-dystrophin plasmid (pRSVDy-B) into mdx 4cv diaphragm muscle fibres.7,8 In that study, expression of dystrophin in approximately 17% of diaphragm muscle fibres resulted in significant protection of fibres from the damaging effect of repetitive eccentric lengthening contraction, and the recovery of NO synthase at the sarcolemma. Previous studies in transgenic mdx mice have demonstrated that expression of recombinant dystrophin at levels as low as 20% of normal endogenous levels at the myofibre membrane can completely prevent the development of the dystrophic phenotype.9,10 If EBV-based plasmid vectors can yield improved efficiency and retention of the dystrophin transgenes at focal muscle sites such as the diaphragm, which impact significantly on the pathophysiology, or in the digits to improve manual dexterity, then their therapeutic use for local gene therapy in DMD patients may be a prospect in the future. In addition, for applications where the focal ectopic muscle expression of heterologous therapeutic transgenes is proposed, the use of EBV-based plasmid vectors with episomal replication and nuclear retention function may likewise prove extremely attractive.

Acknowledgements This work was supported by the UK Medical Research Council, the Leopold Muller Trust and the Muscular Dystrophy Group of Great Britain and Northern Ireland.

References 1 Emery AEH. Duchenne Muscular Dystrophy, 2nd edn. Oxford Medical Publications: Oxford, 1993. 2 Brown SC, Dickson G. Duchenne Muscular Dystrophy. In: Dickson G (ed). Molecular and Cell Biology of Human Gene Therapeutics. Chapman & Hall: London, 1995, pp 261–280. 3 Acsadi G et al. Human dystrophin expression in mdx mice after intra-muscular injection of DNA constructs. Nature 1991; 352: 815–818. 4 Acsadi G et al. Dystrophin expression in muscles of mdx mice after adenovirus-mediated in vivo gene transfer. Hum Gene Ther 1995; 1: 129–140. 5 Dickson G. Gene therapy for Duchenne muscular dystrophy. Chem Indust 1997; 8: 294–297. 6 Dickson G et al. Human dystrophin gene transfer: production and expression of a functional recombinant DNA-based gene. Hum Genet 1991; 88: 53–58. 7 Decrouy A et al. Mini-dystrophin gene transfer in mdx4cv diaphragm muscle fibers increases sarcolemmal stability. Gene Therapy 1997; 4: 401–408. 8 Decrouy A et al. Mini- and full-length dystrophin gene transfer induces the recovery of nitric oxide synthase at the sarcolemma of mdx4cv skeletal muscle fibers. Gene Therapy 1998; 5: 59–64.


9 Phelps SF et al. Prevention of muscular dystrophy by full-length and internally truncated dystrophins. Hum Mol Genet 1995; 4: 1251–1258. 10 Wells DJ et al. Expression of human full-length and minidystrophin in transgenic mdx mice: implications for gene therapy of Duchenne muscular dystrophy. Hum Mol Genet 1995; 4: 1245– 1250. 11 Dunckley MG, Wells DJ, Walsh FS, Dickson G. Direct retroviralmediated transfer of a dystrophin minigene into mdx mouse muscle in vivo. Hum Mol Genet 1993; 2: 717–723. 12 Fassati A et al. Genetic correction of dystrophin deficiency and skeletal muscle remodelling in adult mdx mouse via transplantation of retroviral producer cells. J Clin Invest 1997; 100: 620– 628. 13 Vincent N et al. Long-term correction of mouse dystrophic degeneration by adenovirus-mediated transfer of a minidystrophin gene. Nat Genet 1993; 5: 130–134. 14 Dodds E et al. Lipofection of cultured mouse muscle cells: a direct comparison of Lipofectamine and DOSPER. Gene Therapy 1998; 5: 542–551. 15 Trivedi R, Dickson G. Liposome-mediated gene transfer into normal and dystrophin-deficient mouse myoblasts. J Neurochem 1995; 64: 2230–2238. 16 Yanagihara I, Inui K, Dickson G, Okada S. Expression of fulllength human dystrophin cDNA in MDX mouse by HVJ-liposome injection. Gene Therapy 1996; 3: 549–553. 17 Dickson G. Gene transfer to muscle. Biochem Soc Trans 1996; 24: 514–519. 18 Wells KE et al. Immune responses, not promoter inactivation, are responsible for decreased long-term expression following plasmid gene transfer into skeletal muscle. FEBS Lett 1997; 407: 164–168. 19 Wells DJ et al. Evaluation of plasmid DNA for in vivo gene therapy: factors affecting the number of transfected fibers. J Pharm Sci 1998; 87: 763–768. 20 Margolskee RF. Epstein–Barr virus expression vectors. Curr Topics Microbiol Immunol 1992; 158: 67–95. 21 Yates JL, Guan N. Epstein–Barr virus-derived plasmids replicate only once per cell cycle and are not amplified after entry into cells. J Virol 1991; 65: 483–488. 22 Krysan PJ, Haase SB, Calos MP. Isolation of human sequences that replicate autonomously in human cells. Mol Cell Biol 1989; 9: 1026–1033. 23 Jankelevich S, Kolman JL, Bodnar JW, Miller G. A nuclear matrix attachment region organizes the Epstein–Barr viral plasmid in Raji cells into a single DNA domain. EMBO J 1992; 11: 1165– 1176.

24 Mellon P, Parker R, Gluzman Y, Maniatis T. Identification of DNA sequences required for transcription of the human ␣-1globin gene in a new SV40 host-vector system. Cell 1981; 27: 279–288. 25 Queen C, Baltimore D. Immunogloblin gene-transcription is activated by downstream sequence elements. Cell 1983; 33: 341–348. 26 Mecsas J, Sugden B. Replication of plasmids derived from bovine papilloma virus type 1 and Epstein–Barr virus in cells in culture. Ann Rev Cell Biol 1987; 3: 87–108. 27 Dyrks T et al. Generation of bA4 from the amyloid protein precursor and fragments thereof. FEBS Lett 1993; 335: 89–93. 28 Whitesell L, Rosolen A, Neckers LM. Episome-generated N-myc antisense RNA restricts the differentiation potential of primitive neuroectodermal cell lines. Mol Cell Biol 1991; 11: 1360–1371. 29 Pan LC, Margolskee RF, Blau HM. Cloning muscle isoforms of neural cell adhesion molecule using an episomal shuttle vector. Somat Cell Mol Genet 1992; 18: 163–177. 30 Deiss LP, Kimchi A. A genetic tool used to identify thioredoxin as a mediator of growth inhibitory signal. Science 1991; 252: 117–120. 31 Trojan J et al. Loss of tumorigenicity of rat glioblastoma directed by episome-based antisense cDNA transcription of insulin-like growth factor 1. Proc Natl Acad Sci USA 1992; 89: 4874–4878. 32 Lindenboim L, Anderson D, Stein R. The use of Epstein–Barr virus-based shuttle vectors in rat PC12 cells. Cell Mol Neurobiol 1997; 17: 119–127. 33 Krysan PJ, Calos MP. Epstein–Barr virus-based vectors that replicate in rodent cells. Gene 1993; 136: 137–143. 34 Wohlgemuth JG et al. Long-term expression from autonomously replicating vectors in mammalian cells. Gene Therapy 1996; 3: 503–512. 35 Belharz MW et al. Quantitation of muscle precurser cell activity in skeletal muscle by northern blot analysis of MyoD and myogenin expression: application to dystrophic (mdx) mouse muscle. Mol Cell Neurosci 1992; 3: 326–331. 36 DiMario JX, Uzman A, Strohman RC. Fiber regeneration is not persistent in dystrophic (mdx) mouse skeletal muscle. Dev Biol 1991; 48: 314–321. 37 Yaffe D, Saxel O. Serial passaging and differentiation of myogenic cells isolated from dystrophic mouse muscle. Nature 1997; 270: 725–727. 38 Sherratt TG, Vulliamy T, Strong PN. Evolutionary conservation of the dystrophin central rod domain. Biochem J 1992; 287: 755–759.

1335