Dystrophin expression in the mdx mouse after localised and systemic administration of a morpholino antisense oligonucleotide

THE JOURNAL OF GENE MEDICINE R E S E A R C H A R T I C L EJ Gene Med 2006; 8: 207–216.Published online 14 November 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jgm.838

Dystrophin expression in the mdx mouse afterlocalised and systemic administration of amorpholino antisense oligonucleotide

Susan Fletcher*Kaite HoneymanAbbie M. FallPenny L. HardingRussell D. JohnsenSteve D. Wilton

Experimental Molecular MedicineGroup, Centre for Neuromuscular andNeurological Disorders, University ofWestern Australia, Nedlands, Perth,Western Australia, 6097

*Correspondence to: Susan Fletcher,Experimental Molecular MedicineGroup, Centre for Neuromuscularand Neurological Disorders,4th Floor, ‘A’ Block, QE II MedicalCentre, Verdun St, Nedlands, Perth,Australia 6009.E-mail: [email protected]

Received: 12 May 2005Revised: 2 July 2005Accepted: 28 July 2005

Abstract

Background Duchenne and Becker muscular dystrophies are allelicdisorders arising from mutations in the dystrophin gene. Duchenne musculardystrophy is characterised by an absence of functional protein, whileBecker muscular dystrophy is usually caused by in-frame deletions allowingsynthesis of some functional protein. Treatment options are limited, andwe are investigating the potential of transcript manipulation to overcomedisease-causing mutations. Antisense oligonucleotides have been used toinduce specific exon removal during processing of the dystrophin primarytranscript and thereby by-pass protein-truncating mutations. The antisenseoligonucleotide chemistry most widely used to alter pre-mRNA processing is2′-O-methyl-modified bases on a phosphorothioate backbone.

Methods The present studies evaluate 2′-O-methylphosphorothioate, pep-tide nucleic acid and morpholino antisense oligonucleotides in the mdx mousemodel of muscular dystrophy, which has a nonsense mutation in exon 23 ofthe dystrophin gene.

Results We demonstrate dystrophin expression in mdx mouse tissues afterlocalised and systemic delivery of a morpholino antisense oligonucleotidedesigned to target the dystrophin exon 23 donor splice site.

Conclusions The stability of the morpholino structural type, and the factthat it can be delivered to muscle in the absence of a delivery reagent, renderthis compound eminently suitable for consideration for therapeutic exonskipping to address dystrophin mutations. Copyright 2005 John Wiley &Sons, Ltd.

Keywords exon skipping; antisense oligonucleotide chemistries; dystrophinexpression in vivo; mdx mouse; systemic delivery

Introduction

Duchenne muscular dystrophy (DMD) is the most common, serious formof muscular dystrophy, arising from mutations in the dystrophin gene thatpreclude the synthesis of functional protein (for review, see [1,2]). Themilder allelic disorder, Becker muscular dystrophy (BMD), is usually causedby in-frame deletions in the dystrophin gene. In general, there is a correlationbetween the severity of the phenotype and the effect of the deletion on thereading frame. Deletions that disrupt the reading frame result in a severephenotype (DMD), while in-frame deletions are associated with the milder

Copyright 2005 John Wiley & Sons, Ltd.

208 S. Fletcher et al.

disease [3]. In some cases, BMD may only be diagnosedlate in life, clearly indicating that segments of the protein,particularly in the rod domain, are not required fornear-normal dystrophin function [4,5]. Exceptions to thereading frame rule, in particular, deletions involvingexons 3 to 7 [6], and certain point mutations [7,8]are associated with a milder than expected phenotype.Alternative splicing to produce in-frame transcripts hasbeen identified as one mechanism responsible for partiallyovercoming dystrophin gene lesions [9].

Antisense oligonucleotides (AOs) have been used tomanipulate splicing of the dystrophin pre-mRNA to inducespecific exon skipping and thereby eliminate disease-causing mutations [10–16]. The AO chemistry mostwidely used to date to exclude exons from the dystrophingene transcript has been 2′-O-methyl (2OMe)-modifiedbases on a phosphorothioate backbone [12–16]. Severalother chemistries have, and are being evaluated, includingoligodeoxynucleotides (ODNs) [17], locked nucleic acids(LNAs) [18], ethoxy nucleic acids (ENAs) [19], as well assome of the third-generation AOs, such as the unchargedpeptide nucleic acids (PNAs) [18] and morpholinocompounds [18,20]. Advantages and limitations of eachof these chemistries have been reported. The efficacyof some compounds with high nuclease resistance, andhence extended biological persistence, can be limited bypoor cellular uptake. Although LNAs induced efficientexon skipping, they showed a potential lack of specificitydue to exceptionally strong annealing of the LNA to thetarget [18]. There are also conflicting reports regardingthe efficiency of exon skipping induced by morpholinoAOs [18,20].

We have previously reported that a morpholino AOappeared superior to a 2OMePS AO in vitro [20] in themdx mouse model [21] of muscular dystrophy, whichhas a nonsense mutation in exon 23 of the dystrophingene [22]. In these experiments, the morpholino AOwas annealed to a sense-strand leash to permit lipoplexformation, and dystrophin exon 23 skipping could bedetected in vitro after transfection with as little as 5 nMlipoplex. The morpholino AO in the absence of deliveryreagent induced very low levels of dystrophin expressionafter in vitro transfection at 10 µM [20]. We reportedpreviously that 2OMePS AOs were able to restore a lowlevel of dystrophin, in vitro and in vivo, when deliveredas a lipoplex [12], while later studies using the blockco-polymer F127 as a delivery reagent demonstratedimproved dystrophin expression [14].

In this report we describe in vivo studies comparingboth complexed and uncomplexed morpholino and2OMePS AOs of identical sequence. PNAs directedat splice site coordinates shown to be amenable toinduced exon skipping [13] were also assessed. AOswere delivered with and without the block co-polymerF127 and a sense leash. The uptake of double-strandednucleic acids has been reported to be more efficient thanthat of single-stranded antisense oligonucleotides [23]in vitro. The morpholino AO induced exon skipping afterlocalised and systemic delivery in mdx mice in the absence

of a delivery reagent, whereas this same preparation hadpreviously appeared essentially ineffective in vitro [20].The efficacy of the uncomplexed morpholino AO in vivomay confer a substantial advantage when applying suchcompounds in a clinical setting. Although transfection intocultured cells is a critical step in optimising the design ofAOs, ultimately, the efficacy of such compounds can onlybe evaluated in a relevant animal model.

Materials and methods

Design and synthesis of antisenseoligonucleotides

The 2′-O-methylphosphorothioate antisense oligonu-cleotide, designated M23D(+7–18), was initially pre-pared on an Expedite 8909 nucleic acid synthesiser(Applied Biosystems) using the 1 µmol thioate synthe-sis protocol. Gram quantities were subsequently obtainedfrom Avecia (Grangemouth, Scotland). Oligodeoxyri-bonucleotide sense-strand leashes with phosphorothioate(PS) wings and a phosphodiester core were synthe-sised on the Expedite 8909 nucleic acid synthesiser. β-Cyanoethylphosphoramidites and support columns werepurchased from Glen Research (Stirling, VA, USA). Themorpholino AO, M23D(+7–18), was obtained from Gene-Tools (Philomath, OR, USA) and the peptide nucleic acids(PNAs) were supplied by Applied Biosystems (Foster City,CA, USA).

To facilitate a direct comparison between the exon-skipping efficacy of the morpholino and 2OMePSchemistries, both AOs were designed to anneal to the last7 bases of exon 23 and the first 18 bases of intron 23 [20].Details of the AO sequences are shown in Table 1. PNAscould not be designed to anneal to the same coordinatesdue to internal self-complementarity as specified by themanufacturer.

Annealing of AO to leash andtransfection complexes

Morpholino M23D(+7–18), 2OMePS M23D(+7–18)and PNA M23D(−2–16) and M23D(−2–18) AOs were

Table 1. Oligonucleotide sequences and chemistry. Antisenseoligonucleotides are shown in bold in the 5′ - 3′ ori-entation and sense leashes are shown 3′-5′, -OO- = linker,K = lysine, and chemistries are indicated according to thefollowing key: phosphodiester (PO) = lowercase phosphoroth-ioate (PS) = UPPERCASE. Only leash bases complementary toM23D(+7–18) are underlined

Nomenclature Chemistry Oligonucleotide sequence

M23D(+7–18) Morpholino ggccaaacctcggcttacctgaaatM23D(+7–18) 2OMePS ggccaaacctcggcttacctgaaatLeash 3 PS/PO/PS TAGTGccggtttggagccgaatGTTAGpM23D(−2–16) PNA ccaaacctcggcttapM23D(−2–16)4K PNA ccaaacctcggctta-oo-KKKKpM23D(−2–18)4K PNA ggccaaacctcggctta-oo-KKKKLeash 8 PS/PO/PS AACACGttccggtttggagccgaaTGGAA-

CAACC

Copyright 2005 John Wiley & Sons, Ltd. J Gene Med 2006; 8: 207–216.

Dystrophin Expression in the mdx Mouse 209

annealed to a sense-strand leash according to the protocolof Braasch and Corey [24]. Block co-polymer transfectionmixes were prepared by combining F127 (Sigma, MI,USA) and AOs in normal saline as described previously[14]. Morpholino and 2OMePS M23D(+7–18) AOs wereannealed to Leash 3 and PNA M23D(−2–16) andM23D(−2–18) were annealed to Leash 8. Sequences areshown in Table 1.

In vivo AO injections

All AOs were prepared in normal saline and also as aduplex with a sense-strand leash [20] and the blockco-polymer, F127 [14].

Intramuscular injection of AOpreparations

One tibialis anterior muscle of each experimental mdxmouse was injected with a 10 µl dose of the AOpreparations and the contralateral muscle was injectedwith either saline or the delivery reagent. The animalswere sacrificed at various time points after injection, themuscles were removed and frozen in isopentane cooledin liquid nitrogen, before being sectioned and preparedfor RNA, protein studies and immunofluorescencestudies.

Systemic delivery of AOs

Morpholino M23D(+7–18) (10 mg/ml) in normal salinewas delivered to neonatal mdx mice by the intraperi-toneal route, at a dosage of 25 mg/kg either assingle or multiple injections. Multiple injections ofmorpholino M23D(+7–18) were given to neonatalmdx mice by the intraperitoneal route, at a dose of25 mg/kg/day for 7 days. Thereafter, the injection fre-quency was reduced to three times a week for afurther 2 weeks. Six weeks after the initial injection,mice were euthanased, and tissues removed for anal-ysis by reverse transcription polymerase chain reaction(RT-PCR), Western blotting and immunofluorescencestudies.

RNA preparation and RT-PCR analysis

RNA was extracted from 2–3 mg of sections from frozentissue blocks, using Trizol (Invitrogen), according tothe manufacturer’s protocol. RT-PCR was performedacross dystrophin exons 20–26 as previously described[12,25]. Amplification products were fractionated on2% agarose gels, stained with ethidium bromide andthe images captured by a Chemi-Smart 3000 geldocumentation system (Vilber Lourmat, Marne La Vallee,France).

Protein extraction and Westernblotting

Protein extracts were prepared by adding 120 µl oftreatment buffer consisting of 125 mM Tris/HCl, pH6.8, 4% sodium dodecyl sulfate (SDS), 40% glycerol,0.5 mM phenylmethylsulfonyl fluoride (PMSF), 50 mMdithiothreitol, bromophenol blue and protease inhibitorcocktail (Sigma) to 5 mg mouse muscle cryostat sections.Samples were then vortexed briefly, sonicated for2 s four to eight times, and then heated at 95 ◦Cfor 5 min, before being fractionated at 16 ◦C on a3–10% SDS gradient gel at pH 8.8 with a 3%stacking gel, pH 6.8. Proteins were transferred fromthe gel to Hybond nitrocellulose (Amersham Biosciences,Castle Hill, Australia) overnight at 18 ◦C, at 290 mA.Dystrophin was visualised using NCL-DYS2 monoclonalanti-dystrophin (Novocastra, Newcastle-upon-Tyne, UK)at a dilution of 1 : 100 for 2 h at room temperature withsubsequent detection using the Western Breeze proteindetection kit (Invitrogen). Images were captured directlyby a Chemi-Smart 3000 gel documentation system usingChemi-Capt software for image acquisition and Bio-1Dsoftware for image analysis (Vilber Lourmat).

Tissue preparation and dystrophinimmunofluorescence

Muscle samples were snap-frozen in isopentane, pre-cooled in liquid nitrogen. Dystrophin was detected in6 µm unfixed cryostat sections using the Novocastra NCL-DYS2 monoclonal antibody that reacts strongly withthe C-terminus of dystrophin. Immunofluorescence wasperformed using the Zenon Alexa Fluor 488 labelling kit(Invitrogen), according to the protocol recommended bythe manufacturer, but omitting the initial fixation step.The primary antibody was used at a dilution of 1 : 10with a molar ratio of 4.5 : 1. Where indicated, sectionswere also stained with Hoechst (Sigma) diluted 1 : 10 000to visualise the nuclei. Sections were viewed with anOlympus IX 70 inverted microscope and the imagescaptured on an Olympus DP 70 digital camera.

Results

Intramuscular injection of AOcomplexes

Total RNA was prepared from mdx mouse tibialis anteriormuscle (2–3 mg). Typically, 4–8 mice were injected at6 weeks of age, with AO preparations and vehicle asindicated and analysed 2 weeks later. Representative datais shown in Figures 1 and 2. The full-length transcript isrepresented by an amplicon of 903 bp and is present in allRT-PCR reactions except the PCR negative control. The688 bp product represents the transcript missing exon 23.



Figure 1. RT-PCR analysis of transcripts from RNA prepared from mdx mouse muscle. (a) Analysis of RNA prepared from mdxmouse tibialis anterior muscle injected with a single dose of 5 or 10 µg of each AO in normal saline. Total RNA was extractedfrom 2–3 mg of tissue and amplified by nested RT-PCR across exons 20–26. (b) Analysis of RNA prepared from mdx mouse tibialisanterior, 2 weeks after a single injection of 2 µg of each AO prepared with or without leash and F127. The full-length transcriptis represented by a 903 bp product amplified by nested RT-PCR between dystrophin exons 20–26. The 688 bp product representsthe exon 23-deleted transcript. The outer lanes contain a 100 bp ladder. (c) Western blot analysis of dystrophin expression inAO and sham-injected mdx mouse tibialis anterior. Sample loading was normalised for myosin heavy chain expression (data notshown) and an extract from muscle from a C57 Bl/10 adult mouse was included for comparison. (d) Immunofluorescent staining ofdystrophin on unfixed 6 µm cryosections from AO and sham-injected mdx mouse muscle and from normal C57Bl/10 mouse muscle,using Novocastra NCL-DYS2 monoclonal antibody and Zenon Alexafluor 488. Sections were prepared from muscle, 2 weeks after asingle injection of each AO alone and in combination with F127, with/without leash

Figure 1 shows relative levels of dystrophin exon 23skipping 2 weeks after 5 and 10 µg of uncomplexed AOwere injected into tibialis anterior muscles of mdx mice.The 688 bp product was detected at low levels in tissuetreated with 10 µg of the 2OMePS AO. The shortenedtranscript appeared as a major product in samples from

muscle injected with 5 and 10 µg of the morpholinoM23D(+07–18). No shortened transcript was detected inmuscle treated with the two PNA-4Ks or the lower doseof 2OMePS M23D(+07–18).

In subsequent experiments, AOs (2 µg) were injectedinto the tibialis anterior muscles of mdx mice (6 weeks



Figure 2. RT-PCR (a) and Western blot (b) of Dystrophin expression (c) after injection of 2, 5 and 10 µg doses of morpholinoM23D(+7–18) into mdx mouse tibialis anterior at 11 days of age. The mice were sacrificed 4 weeks later. Dystrophin expressionwas detected with Novocastra NCL-DYS2 monoclonal antibody and Zenon Alexafluor 488 on cryosections of muscle from mdx miceinjected with morpholino M23D(+7–18) at 11 days of age (upper panel) and 36 weeks of age (lower panel) and sacrificed 2 weekslater. Normal C57Bl/10 and sham-injected mdx muscle sections are included for comparison

old) and as duplexes (annealed to sense leash) with theblock co-polymer, F127. The shortened transcript wasonly detected in muscle which had been injected withmorpholino M23D(+07–18) preparations (Figure 1b).Western blot analysis of muscle extracts demonstratedsubstantial levels of dystrophin in muscle injectedwith 5 and 10 µg of morpholino M23D(+07–18)in saline. Dystrophin was not found in the sham-injected mdx muscle extract or the muscles injectedwith the 2OMePS M23D(+07–18) or the PNA-4K(Figure 1c).

2OMePS M23D(+07–18) alone or with F127, with andwithout leash, induced low levels of immunofluorescentstaining for dystrophin at the periphery of musclefibres on unfixed cryosections (Figure 1d). MorpholinoM23D(+07–18) in saline induced more widespread and

intense dystrophin expression compared to the otherpreparations.

Evaluation of morpholinoM23D(+07–18) dosage

As shown in Figures 2a (RT-PCR) and 2b (Westernblot), a dose of 5 µg morpholino M23D(+7–18) insaline resulted in more abundant shortened transcriptand dystrophin than the 2 and 10 µg treatments(n = 4). Immunofluorescent staining for dystrophin oncryosections from injected mdx mouse muscle (Figure 2c)supports this observation. Injection of 36-week-old mdxmice (n = 3) with 5 and 10 µg doses of the morpholino AOalso induced widespread dystrophin expression 2 weekslater, as seen in Figure 2c.



Figure 3. Immunofluorescent staining of dystrophin on cryosections from mdx mouse tibialis anterior, 6 weeks after a singleinjection of 5 µg of morpholino M23D(+7–18) in normal saline at 11 days of age. Dystrophin was detected with NovocastraNCL-DYS2 monoclonal antibody and Zenon Alexafluor 488 on the entire muscle section (4 × magnification) (A) and shown at10 × magnification and counterstained with Hoechst to visualise nuclei (B). An enlarged view (C) shows peripheral (blue arrow)and centrally (red arrow) located nuclei on the muscle fibres

Effect of morpholino M23D(+7–18)injection on mdx mouse muscle

Near-normal dystrophin staining in 36% of muscle fibreswas observed after a 5 µg dose of the morpholinoM23D(+7–18) in saline was injected into the tibialisanterior muscle of 11-day-old mdx mice (Figure 3A).Counterstaining with Hoescht (Sigma) demonstrated areduction in the proportion of centrally nucleated fibres(8%) compared to adjacent dystrophin negative tissue(22%), 6 weeks after injection. Mononuclear cells areprevalent in the dystrophin-negative tissue, while theirincidence is reduced in adjacent regions where dystrophinexpression has been restored (Figure 3B). Figure 3Cshows an enlarged view of muscle fibres demonstratingcentrally and peripherally located nuclei.

Systemic delivery of morpholinoM23D(+7–18) to neonatal mdx mice

Neonatal mdx mice were given either single or multipleintraperitoneal injections of morpholino M23D(+7–18)in normal saline, at a dosage of 25 mg/kg, and sacrificedat 7 weeks of age. The induced shortened transcript wasdetected at low levels in tibialis anterior, quadricepsfemoris, triceps brachialis and diaphragm in mdx micegiven multiple intraperitoneal injections of morpholinoM23D(+7–18) (Figure 4a). The RT-PCR product ofapproximately 750 bp represents spontaneous skippingof exon 21 and has been reported previously [12].Dystrophin was detected by Western blotting in tibialisanterior and diaphragm from these mice (Figure 4b) butnot in tissues from mice given only a single injection (datanot shown). Immunofluorescent staining for dystrophinon tissue sections demonstrated patches of dystrophin-positive fibres and some more widespread, low-levelfluorescence in diaphragm, ileum and tibialis anterior inmdx mice given both the single and multiple treatments(Figure 4c). Some morpholino M23D(+7–18) systemicinjections resulted in marked dystrophin expressionaround blood vessels.

Discussion

A number of approaches, including gene or cell replace-ment, homologous gene up-regulation and nonsensemutation suppression, are being investigated to addressthe dystrophin deficiency in DMD (for review, see[26,27]). To date, the most significant improvementsin DMD prognosis have come from steroid treatmentsand nocturnal mechanically assisted ventilation [28]. It isnot unreasonable to assume that, if reduced respiratorystress can substantially improve the life of a DMD patient,benefits from any treatment that partially restores dys-trophin expression are likely to be significant, particularlywhen used in conjunction with the current managementpractises.

It has been estimated that >75% of DMD patientscould benefit from exon removal to overcome dystrophinmutations [29]. The efficient application of specific exonskipping to by-pass dystrophin mutations could resultin a shortened dystrophin of at least partial function.Targeted exon skipping will not be applicable to all DMDpatients, in particular those affected by large genomicdeletions or the loss of essential coding domains. Althoughgenomic deletions including large or essential proteincoding domains should respond to AO application atthe RNA level, the resultant protein could be severelycompromised with respect to function.

The utility of AO-induced exon skipping in restoringdystrophin expression in experimental systems has beendemonstrated [11–16]. The current challenge is toachieve sustained dystrophin expression, with minimaladverse effects, in most, if not all, tissues affected bythe absence of dystrophin. Ideally, AOs for induced exonskipping should be designed to the most amenable targetmotif, show efficacy at low doses, have sustained actionand preferably be easily delivered as a simple formulation.

The majority of dystrophin exon-skipping studies todate have used 2OMePS AOs to redirect pre-mRNAprocessing [10,12–15,30,31]. We have shown previouslythat a morpholino AO appeared more effective in inducingexon 23 skipping after intramuscular injection than theoptimal 2OMePS AO available at that time [20]. In this



Figure 4. Analysis of dystrophin expression in tissue samples from an mdx mouse given multiple intraperitoneal injections ofmorpholino M23D(+7–18) (25 mg/kg). (a) Total RNA was extracted from 2–3 mg of tissue and amplified by nested RT-PCR acrossexons 20–26. The sizes of the intact and induced shortened transcripts are indicated. (b) Western blot of extracts of tissues froma treated mdx mouse. (c) Immunofluorescent staining of dystrophin on unfixed cryostat sections of mouse tissue using NovocastraNCL-DYS2 monoclonal antibody and Zenon Alexafluor 488. Neonatal mdx mice received either single or multiple intraperitonealinjections of the morpholino M23D(+7–18) (25 mg/kg) and were sacrificed at 7 weeks of age. Tissue sections from normalC57Bl/10 and untreated mdx mice are included for comparison. Red arrows indicate representative fibres showing dystrophinstaining. The blue arrow indicates a putative single revertant fibre in cardiac muscle

study we have further explored the potential of themorpholino chemistry and evaluated delivery of 2OMePSand morpholino AOs for exon skipping in vivo. Thenucleotide sequences of the 2OMePS and morpholinoAOs were identical.

We previously demonstrated that the efficient in vitrouptake of a morpholino AO was dependent on duplex

formation with a charged sense-strand leash and deliveryas a lipoplex [20]. It has recently been reported that theuptake of double-stranded nucleic acids is more efficientthan that of single-stranded antisense oligonucleotides[21] in vitro. Our previous in vivo studies, in which themorpholino AO duplex : lipoplex was administered tomdx mice by intramuscular injection, showed substantial



amounts of induced dystrophin [20]. These studies havenow been extended and we report a comparison betweendifferent AO chemistries delivered alone, or as duplexeswith and without a compound to enhance delivery.While previous in vivo studies were undertaken usinga commercially available cationic liposome [20], it hassince been shown that the block co-polymer F127 is aneffective delivery vehicle for a 2OMePS AO [14,15].

We tested the different AO chemistries in vivo andobserved that the morpholino AO induced substantialshortened transcript 2 weeks after single intramuscularinjections (2, 5 and 10 µg). The intact and inducedtranscripts were present in almost equal amounts,whereas the 2OMePS AO only generated traces of theshortened transcripts at the higher dose. This trendwas confirmed by Western blot and immunofluorescenceanalysis, which showed substantial dystrophin expressioninduced only by the morpholino M23D(+7–18). Thedose of the morpholino AO was evaluated and foundto be optimal at 5 µg per 10 µl intramuscular injection.Morpholino AOs are reputedly non-toxic [32] and haveminimal non-antisense effects [33], and, although therewere no overt signs of adverse reaction 2–4 weeks afterthe higher dose of 10 µg was administered to mice byintramuscular injection, induced dystrophin expressionwas lower, as demonstrated by RT-PCR, Western blottingand immunofluorescent staining.

When 2 µg of AO, with or without sense-strand leashand F127, was injected into the tibialis anterior muscle ofmdx mice, only the morpholino AO induced the shorteneddystrophin transcript when assessed after 2 weeks. Thelack of dystrophin transcripts missing exon 23 in muscleinjected with the 2OMePS AO is inconsistent with otherreports [14] and suggests that the 2 µg dose may bebelow the threshold for induced exon skipping. Data fromRNA and immunofluorescent studies presented here showthat the morpholino AO was superior to the 2OMePSAO and the PNA-4Ks at inducing dystrophin expression.It has been reported that a 2OMePS AO directed toexon 46 in a humanised mouse model induced maximalexon 46 skipping 1–2 weeks after administration [34].Shortened transcripts were reported up to 28 days afterinjection, but it should be noted that the maximalskipping corresponded to only 3% of the total transcriptsand was achieved at much higher AO doses, usingpolyethylenimine as a delivery reagent [34].

The ability of the uncomplexed morpholino AO toinduce exon skipping was unexpectedly high in vivo,whereas this same preparation had previously appearedineffective in vitro. The efficient uptake of the uncom-plexed morpholino AO in vivo may confer a substantialadvantage when applying these compounds in the clinicalsetting. Although transfection into cultured cells is criticalin optimising the design of AOs to induce exon skipping,the ultimate efficacy of AOs can only be evaluated in arelevant and appropriate animal model.

It is generally presumed that if AO-induced exonskipping is to provide the greatest benefit to DMD patients,it will be necessary to commence treatment before

irreversible muscle loss has occurred. We injected theleft tibialis anterior muscles of a litter of 11-day-old mdxmice with the morpholino AO in saline, and cryosectionswere stained to detect dystrophin and nuclear localisation,6 weeks later. Extensive areas of the muscle sectionshowed near-normal appearance and dystrophin staining.In the region where dystrophin was expressed, the numberof mononuclear cells and the degree of central nucleationwere reduced when compared to the adjacent dystrophin-negative tissue.

Although intramuscular injection of AOs is usefulin evaluating the efficacy of compounds in vivo, anytreatment for DMD will need to be effective aftersystemic delivery. Because skeletal muscle constitutesabout 30% of the total body mass and because dystrophinisoforms are also expressed in a variety of tissues,DMD can only be ameliorated by systemic dystrophinrestoration. We evaluated intraperitoneal injection ofthe morpholino AO as a potential systemic treatmentregimen. Immunofluorescent staining of sections ofselected tissues from 7-week-old mdx mice, which hadreceived a single intraperitoneal injection at 5 days ofage, revealed the presence of a low level of dystrophinin the tibialis anterior, quadriceps femoris and diaphragm.No dystrophin was detected by Western blotting anddystrophin transcripts excluding exon 23 were notconsistently observed.

Dystrophin expression was subsequently demonstratedby immunofluorescent staining in mdx mouse tibialis ante-rior, ileum, diaphragm, triceps brachialis and quadricepsfemoris after a series of 13 intraperitoneal injections,starting at 2 days of age and continuing over a 3-weekperiod. These animals were sacrificed at the same ageas those that received a single treatment. However, theshortened transcript was detected in limb muscles anddiaphragm and Western blot analysis revealed low levelsof dystrophin in tibialis anterior and a trace amount indiaphragm.

The progression of muscle wasting in DMD is relentless.It would be preferable for AO treatment to commenceas early as possible, prior to extensive muscle damageoccurring and before the muscle architecture is severelydisrupted. Restoration of dystrophin expression in youngmdx mice injected with the morpholino AO, prior to thepeak of muscle degeneration [35], resulted in a reducedpercentage of centrally nucleated fibres, muscle fibresof near-normal appearance and reduced infiltration ofmononuclear cells. This result may reflect that the majorepisode of muscle degeneration, which normally occurs inthis disease model, has been minimised by the presenceof dystrophin at the sarcolemma of the majority of musclefibres in the treatment area and strongly suggests thatthe induced dystrophin is functional. Persistence of theshortened transcript for at least 4 weeks, and the presenceof substantial sarcolemmal dystrophin 6–10 weeks after asingle intramuscular injection of the morpholino AO, areencouraging. However, additional long-term studies arerequired to establish the level, extent and duration of thebenefit and to design a treatment regimen that achieves



sufficient dystrophin expression to prevent muscle lossand preserve muscle function. Systemic administrationof AOs by a number of different routes, includingintravenous, transdermal [36], subcutaneous and oral[37–39], has been reported.

Although amelioration of the dystrophic phenotype byexon skipping has only been reported in the mdx mouse[14,15], dystrophin exon skipping in human cells hasbeen reported [16,25]. It is yet to be demonstrated thatremoval of dystrophin exons can reduce the phenotypeseverity in DMD or BMD patients. Although it is predictedthat many DMD patients could benefit from AO-inducedexon skipping, each mutation will need to be evaluatedindependently. It is probable that benefits from thetreatment will vary substantially between patients andresponses are likely to be dependent on the nature andposition of the mutations and the age at which treatmentcommences.

A major shortcoming of exon removal to treat DMDis that the AOs will need to be administered regularly tofirstly establish, and then maintain, dystrophin expression.In an attempt to overcome this limitation, persistentdystrophin exon 23 skipping in the mdx mouse has beenachieved by the single administration of an AAV vectorexpressing antisense sequences, linked to a modified U7small nuclear RNA [40]. However, as it may be consideredpreferable to avoid the use of viral vectors in the clinic, anAO that has extended biological stability, minimal toxicityand non-antisense effects could be highly desirable. Thestability of the morpholino structural type [13,32,33,41],and the fact that the AO can be delivered to muscle in theabsence of a delivery reagent, render this AO chemistryeminently suitable for consideration for therapeutic exonskipping to address DMD mutations.

Acknowledgements

The authors receive funding from the National Institutes ofHealth (RO1 NS044146-02), the Muscular Dystrophy AssociationUSA (MDA3718), the National Health and Medical ResearchCouncil of Australia (303216), and Aktion Benni & co e.V.,Germany. The technical assistance of Joshua Steinhaus wasinvaluable.

References

1. Emery AE. Clinical and molecular studies in Duchenne musculardystrophy. Prog Clin Biol Res 1989; 306: 15–28.

2. Emery AE. Muscular dystrophy into the new millennium.Neuromuscul Disord 2002; 12: 343–349.

3. Monaco AP, Bertelson CJ, Liechti-Gallati S, et al. An explanationfor the phenotypic differences between patients bearing partialdeletions of the DMD locus. Genomics 1988; 2: 90–95.

4. England SB, Nicholson LV, Johnson MA, et al. Very mildmuscular dystrophy associated with the deletion of 46% ofdystrophin. Nature 1990; 343: 180–182.

5. Heald A, Anderson LV, Bushby KM, et al. Becker musculardystrophy with onset after 60 years. Neurology 1994; 44:2388–2390.

6. Muntoni F, Gobbi P, Sewry C, et al. Deletions in the 5′ regionof dystrophin and resulting phenotypes. J Med Genet 1994; 31:843–847.

7. Shiga N, Takeshima Y, Sakamoto H, et al. Disruption of thesplicing enhancer sequence within exon 27 of the dystrophingene by a nonsense mutation induces partial skipping of theexon and is responsible for Becker muscular dystrophy. J ClinInvest 1997; 100: 2204–2210.

8. Fajkusova L, Lukas Z, Tvrdikova M, et al. Novel dystrophinmutations revealed by analysis of dystrophin mRNA: alternativesplicing suppresses the phenotypic effect of a nonsense mutation.Neuromuscul Disord 2001; 11: 133–138.

9. Chelly J, Gilgenkrantz H, Lambert M, et al. Effect of dystrophingene deletions on mRNA levels and processing in Duchenne andBecker muscular dystrophies. Cell 1990; 63: 1239–1248.

10. Aartsma-Rus A, Bremmer-Bout M, Janson AA, et al. Targetedexon skipping as a potential gene correction therapy forDuchenne muscular dystrophy. Neuromuscul Disord 2002;12(Suppl 1): S71–77.

11. Wilton SD, Lloyd F, Carville K, et al. Specific removal of thenonsense mutation from the mdx dystrophin mRNA usingantisense oligonucleotides. Neuromuscul. Disord. 1999; 9:330–338.

12. Mann CJ, Honeyman K, Cheng AJ, et al. Antisense-induced exonskipping and synthesis of dystrophin in the mdx mouse. ProcNatl Acad Sci U S A 2001; 98: 42–47.

13. Mann CJ, Honeyman K, McClorey G, et al. Improved antisenseoligonucleotide induced exon skipping in the mdx mouse modelof muscular dystrophy. J Gene Med 2002; 4: 644–654.

14. Lu QL, Mann CJ, Lou F, et al. Functional amounts of dystrophinproduced by skipping the mutated exon in the mdx dystrophicmouse. Nat Med 2003; 9: 1009–1014.

15. Lu QL, Rabinowitz A, Chen YC, et al. Systemic delivery ofantisense oligoribonucleotide restores dystrophin expression inbody-wide skeletal muscles. Proc Natl Acad Sci U S A 2005; 102:198–203.

16. van Deutekom JC, Bremmer-Bout M, Janson AA, et al.Antisense-induced exon skipping restores dystrophin expressionin DMD patient derived muscle cells. Hum Mol Genet 2001; 10:1547–1554.

17. Pramono ZA, Takeshima Y, Alimsardjono H, et al. Induction ofexon skipping of the dystrophin transcript in lymphoblastoidcells by transfecting an antisense oligodeoxynucleotidecomplementary to an exon recognition sequence. BiochemBiophys Res Commun 1996; 226: 445–449.

18. Aartsma-Rus A, Kaman WE, Bremmer-Bout M, et al. Compara-tive analysis of antisense oligonucleotide analogs for targetedDMD exon 46 skipping in muscle cells. Gene Ther 2004; 11:1391–1398.

19. Yagi M, Takeshima Y, Surono A, et al. Chimeric RNA and 2′-O,4′-C-ethylene-bridged nucleic acids have stronger activity thanphosphorothioate oligodeoxynucleotides in induction of exon19 skipping in dystrophin mRNA. Oligonucleotides 2004; 14:33–40.

20. Gebski BL, Mann CJ, Fletcher S, et al. Morpholino antisenseoligonucleotide induced dystrophin exon 23 skipping in mdxmouse muscle. Hum Mol Genet 2003; 12: 1801–1811.

21. Bulfield G, Siller WG, Wight PA, et al. X chromosome-linkedmuscular dystrophy (mdx) in the mouse. Proc Natl Acad SciU S A 1984; 81: 1189–1192.

22. Sicinski P, Geng Y, Ryder-Cook AS, et al. The molecular basis ofmuscular dystrophy in the mdx mouse: a point mutation. Science1989; 244: 1578–1580.

23. Astriab-Fisher A, Fisher MH, Juliano R, et al. Increased uptakeof antisense oligonucleotides by delivery as double strandedcomplexes. Biochem Pharmacol 2004; 68: 403–407.

24. Braasch DA, Corey DR. Synthesis, analysis, purification, andintracellular delivery of peptide nucleic acids. Methods 2001;23: 97–107.

25. Errington SJ, Mann CJ, Fletcher S, et al. Target selectionfor antisense oligonucleotide induced exon skipping in thedystrophin gene. J Gene Med 2003; 5: 518–527.

26. Bogdanovich S, Perkins KJ, Krag TO, et al. Therapeutics forDuchenne muscular dystrophy: current approaches and futuredirections. J Mol Med 2004; 82: 102–115.

27. van Deutekom JC, van Ommen GJ. Advances in Duchennemuscular dystrophy gene therapy. Nat Rev Genet 2003; 4:774–783.

28. Eagle M, Baudouin SV, Chandler C, et al. Survival in Duchennemuscular dystrophy: improvements in life expectancy since 1967and the impact of home nocturnal ventilation. NeuromusculDisord 2002; 12: 926–929.



29. Aartsma-Rus A, Janson AA, Kaman WE, et al. Antisense-inducedmultiexon skipping for Duchenne muscular dystrophy makesmore sense. Am J Hum Genet 2004; 74: 83–92.

30. Aartsma-Rus A, Janson AA, Kaman WE, et al. Therapeuticantisense-induced exon skipping in cultured muscle cells fromsix different DMD patients. Hum Mol Genet 2003; 12: 907–914.

31. Dunckley MG, Manoharan M, Villiet P, et al. Modification ofsplicing in the dystrophin gene in cultured Mdx muscle cellsby antisense oligoribonucleotides. Hum Mol Genet 1998; 7:1083–1090.

32. Iversen PL. Phosphorodiamidate morpholino oligomers:favorable properties for sequence-specific gene inactivation. CurrOpin Mol Ther 2001; 3: 235–238.

33. Summerton J. Morpholino antisense oligomers: the case foran RNase H-independent structural type. Biochim Biophys Acta1999; 1489: 141–158.

34. Bremmer-Bout M, Aartsma-Rus A, de Meijer EJ, et al. Targetedexon skipping in transgenic hDMD mice: a model fordirect preclinical screening of human-specific antisenseoligonucleotides. Mol Ther 2004; 10: 232–240.

35. Muntoni F, Mateddu A, Marchei F, et al. Muscular weakness inthe mdx mouse. J Neurol Sci 1993; 120: 71–77.

36. Pannier AK, Arora V, Iversen PL, et al. Transdermal deliveryof phosphorodiamidate morpholino oligomers across hairlessmouse skin. Int J Pharm 2004; 275: 217–226.

37. Raoof AA, Chiu P, Ramtoola Z, et al. Oral bioavailability andmultiple dose tolerability of an antisense oligonucleotide tabletformulated with sodium caprate. J Pharm Sci 2004; 93:1431–1439.

38. Arora V, Knapp DC, Reddy MT, et al. Bioavailability and efficacyof antisense morpholino oligomers targeted to c-myc andcytochrome P-450 3A2 following oral administration in rats.J Pharm Sci 2002; 91: 1009–1018.

39. Arora V, Devi GR, Iversen PL. Neutrally charged phospho-rodiamidate morpholino antisense oligomers: uptake, effi-cacy and pharmacokinetics. Curr Pharm Biotechnol 2004; 5:431–439.

40. Goyenvalle A, Vulin A, Fougerousse F, et al. Rescue ofdystrophic muscle through U7 snRNA-mediated exon skipping.Science 2004; 306: 1796–1799.

41. Hudziak RM, Barofsky E, Barofsky DF, et al. Resistanceof morpholino phosphorodiamidate oligomers to enzymaticdegradation. Antisense Nucleic Acid Drug Dev 1996; 6:267–272.


Documents

Dystrophin expression in the mdx mouse after localised and systemic administration of a morpholino antisense oligonucleotide