Mol{:cul::ir ther.ipy: the• }ournal of thr: ..C.:1112cica,
`uur - r;0n~'ra1 Co iie ~·.t.icn
`i,'. 01 :J7o r-ri
`~•;-1
`'-' , r.r). - \r.ug . 2(; 0:J. 1
`'' ·• r-,
`•'1 I '
`1
`' ·
`2nu4 .. oa -.2J
`
`PROPERTY OF THE
`NATIONAL
`LIBRARY OF
`MEDICINE
`
`' ~
`
`\
`k
`
`\
`
`,
`
`_,;(~~.~.; '
`'·
`
`111
`
`·~
`
`·.
`
`

`

`Volume 10, Number 2, August 2004
`
`~ 2004 The AmcrlCi'lrl Society of Gene
`Therapy. All rights reserved.
`This jour11.il Jlld the lruJivldual conulhutlc~ns
`cont,1lncd in it iHI! prolech.>d under copyn9ht by
`l hc Amr.1lc;1n Society of Gene Thcr,,py ,1n<J_ the
`loltowln9 h!1m,; and conditions .-ipply lo their use:
`Photocopyhl!J. Single photocopies of single ,1rll(cid:173)
`c.lcs ni.1y be m,,dc for pcrwn;il t~~e as allowed by
`nntion;il copyright laws. Permission of the Pub·
`llsher and paymc1H ol a k"C .ue requ.l1ed for ,111
`o!her pl1otocopy!n9, including multiple or
`))'\leni.1tlc copyln<J, copying for advertising or
`promolion,11 purposes, res.1le, and all forms ol
`document delivery. Spec.ial rates .ue av.1H,lble for
`etlucatlooal institutions that wish to m:ike pho•
`tocopics fOf non-prolit c<luc,11ional classroom us.e.
`Permlsski11s rn,1y be sought dire,c.Uy from El\cvier
`Glob,11 Ri<Jhls Dcp.u1mcnt, P.O. Box 800,
`Oxfo11J OX5 IDX, UI<; phone: (t-4-1) 18~S
`843830; fax: (t-4-1) 1865 853JJ3; e-mail:
`pc1mi~slon~ ISl.'Vit.'f.CO.uk. You m.1y ,llM> conl<'lCl
`Global RlghU directly Uuou!Jh [lsevid s honlt!
`pa!Jc (htti>://www.elSt'Vier.com/horn.er~ag,;,/JuP(cid:173)
`fXmn by ~lectlng "Obtaining PcnniSSiO/lS. In
`111'! USA, uwrs m,l)' ckoar pcnnissiom nnd m.ikC!
`1f 1J'~~~~?::{~~i1~.c~~~~~~ MA
`
`1,~
`
`:-~
`~ ~~~:
`Ol923, USA; phone: (?78) 750-8'100; f.:uc: (9_/8)
`750--17-1-1; and In th<! UK 1.hrough ti~ ~opyn!.}ht
`Licensing Agency R,1pld Clcar.ince Servtce
`(CLARCS) 90 Tottcnh;11n Court Ro.id, London
`WlP OU~ 'u K; phone: +·4-1 (0)20 7631 5555; f,uc:
`+4-1 (0)20 7631 5500. Other counirlcs m.iy h;we
`a loc,11 r1?pr0<Jrnphic rights ,19ency for paymelltS.
`Ocrlv,,tlve Works. Subsc,lbers_ m,1y reproduce
`lclbles of cor,tents or prep,m? hs~s of "'.tlclt!s
`.
`including ,1bstr.1clS for internal orcul,n10~ w,ll~lll
`their inslitutions. Petmission of the Pu~ishr IS
`r~~~i~~~r~~rp~~~::s:nd~t~~~:~~~l~~~s: 1: t ,e
`requited for ,111 other derivative works, Includ(cid:173)
`ing compilations and trnnsla1ions.
`Eleclronlc Storage or Us.l9e. Permls~on of the
`Publisher Is required lo sto,c or use ei...:tronl~ally
`any malcrlal cont,1tne<J in this Journ.11, 11,clud•~ged
`,my article or part of an articll?. E11.cept as ootl,n
`~bove no pa,t of this publication may be reJ>:fO•
`duc.ed: i.lored in a rettil.-Vill system, or ~rarnn11t1ed
`
`~:~~,~~~~' ~,~r;'~;i~;~:~!;~;~1
`
`; other•
`wise withoul prior written pt.>rm1ssK>O of the
`Publisht!r. Address permission requests to lhe .
`Elsevier Global Rights Dep.irtmen1, at the m.-ul,
`fax and e-mall addresses no1ed above.
`No~lce No responsibility is assum~ by the Pub·
`lishcr f~r Ml)' ln/ury and/or damnge .'? pcrso~s or
`property i!S a m.iuer of product Hab1hty, neg ·
`genc.c, or otherwise, or from any use O~ opera•
`~~~s0~!~ii~:~1~~~rm~t~~gi ~sr~i~~t;~u~ of
`,,1pid adv,mce1. in the me~ical scie.nces, Ill par~c(cid:173)
`ul.ir, indept.>ndent venficat1on of d1ag~ s an
`drug dosages should be m..1de,
`Although all advenising matcrlal is expected to
`conform to ethical (medical) standards,
`.
`Inclusion In this publlc.itlon does not c~~lst1tute
`a qu,uantee or endorsement of th~ qu,, ,ty or f
`value of such product or of the claims made o
`il by Its ,n,,nufacturer.
`1 S25-0016/04 S30.00
`M..ot: IN lnl IJNIIIO U Al fi OI AMf ~iCA
`
`J\'\OI.ECUI.AR T IIJ-:R,\l'Y
`(ISSN 1525-0016)
`Piibllcatlon lnformiltlon: Molecufar Th~ropy
`(ISSN 1525-0016) is pubfohed blm?ntt,I_Y by
`Academic Press, an imptint of f.!sev1er, 5:ZS B
`Street Suite 1900, Sm1 Oieqo, CA 9210,1-4495,
`USA. for 2004 Volume(s) 9 and 1.0, .12 is~ue,s arc
`scheduled for p ublication. Sub~C11puon prices
`.-ire available upon request from the Publisher,
`ei~~1~1~1s~~~~~~~~~:o~;I~~~::,:~: or
`S~bscriplion orders arc entered for 1.he calendar
`year and arc payab.le In adv.:ir,ce. POSTMASTCR:
`send ch,mges of address to Mo/t?Ciilar I/Jerapy,
`flX?viCr, Customer Support Department, 6277
`Se.i Harbor Drive, Qtl;mdo, Fl 3288/-4900,
`USA. Periodicals posta9e paid at ~ tl,mdo, Fl.
`USA, and at adtlition.il mailing olhces.
`Orders, clalms, and product Inquiries: Ple:ne
`conlacl the Customer Support Dep.irtment at
`the Regfonal Salt.>s Oflice nearest you, New York:
`El~ev!er, PO Box 945, New Yo1k,_ NY 10159-
`0945, USA; phone: (+1) (21~) 6 33 3710 [1~11-
`free number for North American customers.
`1-886--iES-INfO (4 37-46 36)]; f.-.x: (t-l) (21 2)
`633 3680; e-mail: usinfo,fC!9elsevicr.com.
`Amsterdam: Elsevier, PO Box 211, 1000 AE
`Amsterd.im, The Netherlands; phone; (-~ 31) 20
`4853757; fox; (t-31) 20 4853432; ~-mail:
`nllnfo-f@elsevicr.com. Tokyo: El~ev1cr, 9-1 S
`Higashl-Alabu 1-chomt>, Minato-ku, Tokyo
`106-0044, J.ipan; phone: (-•81).0) 5561 5033;
`fax: {,+81) ()) S561 5047; e-mail:
`infot!Yclsevler.co.jµ. Singapore: Elsevier, 3
`Kllliney Road, #08-01 Winsland House I,
`Singapore 239519; phone: _(-1'65) 6349 0200;
`fax: (+65) 6733 15 10; e-mail:
`asi.iinfo'.!i'elsevicr.com.sg. Rio de Janeiro:
`Elsevier, Rua Sele de Setemlvo 11 1 /16 An.d.ir,
`20050-002 Centro, Rio de Janeiro•RJ, Brazil;
`phone: (+SS) (21) 50'J 53·10; fax: (+55) (21) 507
`1991; e-mail: e1sevler@campus.com.br. (Note
`Latin Ame,ica: For orders, d.iims and help desk
`info1maUon, please cont.ict the Reqlo11al S11les
`Of/ice In Nl.'W York as lis.ted nbove.]
`@ The paper used in this publication meets the
`requirements of ANSI/ NISO Z39.48-1992
`{Pe,mancnce of P,,per)
`
`REGULAR ARTICLES (CONT.)
`
`Integration and Long-Tenn Expression in Xenografted Human Glioblastoma Cells
`Using a Plasmid-Based Transposon System ...... . . . .. . . ...... . ... . . .. . .. ...... 260
`John R. Oh/fest, Paul D. Lobitz, Scott G. Perkinson, and David A. Largaespada
`
`Long-term Transgene Expression from Plasmid DNA Gene Therapy Vectors Js
`Negatively Affected by CpG Dinucleotides ....... . . .. . . ........ .. . .. . . ........ . 269
`Bradley L. Hodges, Kristin M. Taylor, Macy F. Joseph, Sarah A. Bourgeois, and Ronald K. Scheule
`
`HLA-A *0201-Restricte<l Cytolytic Responses to the rtTA Transactivator Dominant
`and Cryptic Epitopes Compromise Transgene Expression Induced by the
`Tetracycline on System . . . ......... . . . .... . ....... ..... . . .. . ....... ... ....... 279
`F. Ginhoux, 5, Turbant, D.A. Gross, J. Poupiot, T. Marais, Y. Lone, F.A. Lemonnier, H. Firat, N. Perez,
`0. Danos, and J. Davoust
`
`Therapeutic HER2/Neu DNA Vaccine Inhibits Mouse Tumor Naturally
`Ovcrexpressing Endogenous Neu . . ..... . . ....... . ...... . ...... .. . .. .. . . . . . .... 290
`Chi-Chen Lin, Ching-Wen Chou, Ai-Li Shiau, Cheng-Fen Tu, Tai-Ming Ko, Yi-Ling Chen,
`Bei-Chang Yang, Mi-Hua Tao, and Ming-Derg Lai
`
`Recombinant AAV Viral Vectors Pseudotyped with Viral Capsids from Serotypes
`I, 2, and 5 Display Differential Efficiency and Cell Tropism after Delivery to
`Different Regions of the Central Nervous System ..... .. ...... .. ... ... .. . .. . . . .. 302
`Corinna Burger, Oleg 5. Gorbatyuk, Margaret J. Velardo, Carmen S. Peden, Philip Williams,
`Sergei Zolotukhin, Paul J. Reier, Ronald J. Mandel, and Nicholas Muzyczka
`
`Intradermal Injection of Lentiviral Vectors Corrects Regenerated Human
`Dystrophic Epidermolysis Bullosa Skin Tissue in Vivo . . . . ... . .... . .. .. ... ... , .. 318
`David T. Woodley, Douglas R. Keene, Tom Atha, Yi Huang, Ramin Ram, Noriyuki Kasahara, and
`Mei Chen
`
`Enhanced Repair of the Anterior Cmciate Ligament by in Situ Gene Transfer:
`Evaluation in an in Vitro Model. ... . .. ........ ....... . . .. . . . .... .... . . ........ 327
`Arnulf Pascher, Andre F. Steinert, Glyn D. Palmer, Oliver Betz, Jean-Noel Gouze, Elvire Gauze,
`Carmencita Pilapil, Stephen C. Ghivizzani, Christopher H. Evans, and Martha Meaney Murray
`
`Adenoviral-Mediated Expression of Porphobilinogen Deaminase in Liver Restores
`the Metabolic Defect in a Mouse Model of Acute Intermittent Porphyria ...... .. .. 337
`Annika Johansson, Grzegorz Nowak, Christer Moller, Pontus Blomberg, and Pauline Harper
`
`Effect of Adenovirus Scrotype 5 Fiber and Penton Modifications on In Vivo
`Tropism in Rats .. . .. . ................. , . , ......... .. . ............... .. ...... 344
`Campbell G. Nicol, Delyth Graham, William H. Miller, Stephen J. White, Theodore A. G. Smith,
`Stuart A. Nicklin, Susan C. Stevenson, and Andrew H. Baker
`
`A Novel TARP-Promoter-Based Adenovirus against Hom1one-Dependent and
`Honnone-Refractory Prostate Cancer . ...... . ..... . . . .. . . .. . . . ........ ... .. . ... 355
`Wing-Shing Cheng, Robert Kraaij, Berith Nilsson, Laura van der Wee/, Corrina M.A. de Ridder,
`Thomas 1-/. Totterman, and Magnus Essand
`
`ii
`
`Th is m at;e,ria I w as c,o,pied
`at the NLM a nd may be
`5ubject USCopyri~ht Law s
`
`M ot.ECUI.AR Tl IEIUl'Y Vol. 10, No. 2, August 2004
`Copyright© The American Society of Gene Therapy
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`2004.05.031
`doi: 10.1 016/" · rnlhe.
`
`Targeted Exon Skipping in Transgenic hDMD Mice:
`A Model for Direct Preclinical Screening of
`Human-Specific Antisense Oligonucleotides
`
`Mattie Bremmer-Bout, Annemieke Aartsma-Rus,
`Emile J. de Meijer, Wendy E. Kaman, Anneke A. M. Janson,
`Rolf H. A. M. Vossen, Gert-Jan B. van Ommen,
`Johan T. den Dunnen, and Judith C. T. van Deutekom*
`Center (or H11111a11 awl Clinical Genetics, Leiden University Medical Center, Wassenaarseweg 72, 2333 AL Leiden, Tile Nelllerloorfs
`
`*To wlwm correspondence and reprint' requests should be 111/dressed. Fax: +3 1-71-5276075. E-111ail: de11lckof11<rN11111c.nl.
`
`Available online 2 July 2004
`
`The therapeutic potential of frame-restoring exon skipping by anti sense oligonucleotides (AONs)
`has recently been demonstrated in cultured muscle cells from a series of Duchenne muscular
`dystrophy (DMD) patients. To facilitate clinical application, in vivo studies in animal models are
`required to develop safe and efficient AON-delivery methods. However, since exon skipping is a
`sequence-specific therapy, it is desirable to target the human DMD gene directly. We therefore
`set up human sequence-specific exon skipping in transgenic mice carrying the full-size human
`gene (hDMD). We initially compared the efficiency and toxicity of intramuscular AON injections
`using different delivery reagents in wild-type mice. At a dose of 3.6 nmol AON and using
`polyethylenimine, the skipping levels accumulated up to 3% in the second week postinjection
`and lasted for 4 weeks. We observed a correlation of this long-term effect with the intramuscular
`persistence of the AON. In regenerating myofibers higher efficiencies (up to 9%) could be
`obtained. Finally, using the optimized protocols in hDMD mice, we were able to induce the
`specific skipping of human DMD exons without affecting the endogenous mouse gene. These
`data highlight the high sequence specificity of this therapy and present the hDMD mouse as a
`unique model to optimize human-specific exon skipping in vivo.
`
`Key Words: Duchenne muscular dystrophy, dystrophin, antisense oligonucleotides,
`exon skipping, polyethylenimine, MALDI-TOF, hDMD mice
`
`)
`
`INTRODUCTION
`In most Duchenne muscular dystrophy (DMD) patients, a
`mutation disrupts the open reading frame of the DMD
`transcript and a shortened, unstable dystrophin protein is
`generated [1-4]. As a consequence, DMD patients suffer
`from progressive muscle fiber degeneration that in time
`lethally affects heart and lungs, typically in early adult(cid:173)
`hood. The incidence of milder Becker muscular dystro(cid:173)
`phy (BMD) phenotypes resulting from (large) in-frame
`deletions [5- 7] has led to the idea of specifically modi(cid:173)
`fying a DMD mutation into its nearest in-frame BMD
`counterpart by inducing exon skipping during pre-mRNA
`splicing. In fact, gene therapy studies based on this
`strategy are currently progressing rapidly. Antisense oli(cid:173)
`gonucleotides (AONs) are applied, which bind to specific
`sequences in the pre-mRNA such that particular exon
`
`inclusion signals (i.e., splice sites, intronic branch-point
`sequences, exon-internal splicing enhancer elements,
`and/or secondary structure) are disturbed or become
`inaccessible for splicing factors. Consequently, the exon
`is spliced out along with its flanking intrans. This exon
`skipping restores the open reading frame and accordingly
`allows the synthesis of llMD-like dystrophins that may
`alleviate progression of the disease.
`The promise of AONs as small-molecule drugs for
`DMD has been demonstrated in numerous studies
`reported over the past 6 years [8-20]. Restoration of
`dystrophin expression was achieved in cultured muscle
`cells from a series of DMD patients affected by different
`mutations and in mdx mouse muscle tissue in vivo. By
`applying AONs that were directed at exon-internal
`sequences in predicted, partially open secondary pre-
`
`232
`
`Th,is mater ial w.is.copied
`attheN LM a.ndma y h,e
`Siu!bject US Co,pyright La ws
`
`MOI.ECUL,\R TIIEiv\l'Y Vol. 10, No. 2, August 2004
`Copyright <0 The American Society of Gene Therapy
`1525-0016/$30.00
`
`

`

`doi:10.1 0l 6/j.ymthe.2004.05.031
`
`mRNA structures, a large series of human-specific AONs
`has to date been identified with which the skipping of
`over 20 different exons can be induced from the DMD
`transcript [8-10, 17, and unpublished results]. This
`would, in theory, be therapeutic for approximately
`70<¾> of DMD mutations. Purthermore, we have achieved
`the specific, simultaneous skipping of two or more
`exons using a combination of AONs [10]. This novel
`strategy of AON-induced multi-exon skipping not only
`enlarges the therapeutic potential for DMD patients, but
`also may allow the engineering of relatively large llMD(cid:173)
`like deletions associated with known mild phenotypes.
`By encompassing a series of smaller DMD mutations,
`this would thus reduce the mutation specificity of this
`therapy.
`Together with the favorable safety aspects of synthetic
`small molecules like AONs, the promising in vitro data
`imply that clinical applications are becoming a realistic
`goal. However, extensive studies in muscle tissue of an
`animal model will be required first to determine and
`optimize the in vivo efficiency and specificity of the
`strategy; to analyze the tissue distribution, toxicity, or
`side effects of AONs; and to develop a more efficient,
`preferably systemic, AON-delivery method. Lu and col(cid:173)
`leagues recently gave convincing in vivo proof of princi(cid:173)
`ple for therapeutic exon skipping in mdx muscle tissue
`[13]. Using the copolymer F127, a widely used inert drug
`vehicle, an AON targeting the 5' splice site consensus
`sequence of mouse exon 23 was injected intramuscular(cid:173)
`ly. Dystrophin expression, resulting from frame-restoring
`exon 23 skipping, was detected in up to 20% of muscle
`fibers. Although this significantly improved muscle
`force, it appeared insufficient to provide a better resis(cid:173)
`tan ce to fatigue. Another local delivery method for
`AONs, based on hyaluronidase treatment and electro(cid:173)
`transfer, has been applied to mdx muscle [21]. Electro(cid:173)
`transfer induces temporary damage to the plasma
`membranes, whereas hyaluronidase hydrolyzes hyalur(cid:173)
`onic acid, a m ajor component of the extracellular matrix
`[22]. The combination of both enhanced AON uptake,
`improved exon 23 skipping efficiencies, and thus re(cid:173)
`stored dystrophin expression in up to 28% of treated
`myotubes. Unfortunately, electrotransfer is accompanied
`by irreversible muscle damage and whole-body treat(cid:173)
`ment seems unappealing. Recently, a transgenic mouse
`model (EGFP-654) was engineered based o n expression
`of the enhanced green fluorescence protein in response
`to a specific AON that prevents aberrant splicing of a
`mutated intron of the human ~-globin gene [23]. This
`model was employed to study the efficiency of different
`chemically modified AON analogs in correcting normal
`splicing. However, since local antisense activity, toxicity,
`and tissue distribution can be visualized directly, this
`model may also be valuable for the development of a
`safe and efficient systemic delivery m ethod for AONs in
`DMD.
`
`While there is another, larger and clinically more
`represen~at~ve,. animal model for DMD, the GRMD dog
`[24], opt1m1zat10n of the AON-based exon-skipping ther(cid:173)
`apy will, for practical and financial reasons, be performed
`mainly in mice. However, the AON approach entails a
`sequence-specific therapy, whereas the human and
`mouse target sequences can be markedly different. It will
`thus be hard to optimize human-specific exon skipping
`using the mouse DMD gene. Therefore, we have estab(cid:173)
`lished the transgenic hDMD mouse, carrying an integrat(cid:173)
`ed copy of the full-length human DMD gene, which we
`present here as a unique model to test directly human(cid:173)
`specific AONs in a mouse experimental background. We
`initially set up targeted exon skipping in mice following
`intramuscular injections of an exon 46-specific AON.
`Different delivery compounds, including polyethyleni(cid:173)
`mine (PEI) and SAINT, were compared, and dosage- and
`time-response effects were assessed in wild-type and mdx
`mice. Using the combination of immunohistochemistry
`and MALDI-TOF mass-spectrometry we further analyzed
`the intramuscular dispersion and stability of the AON at
`different time points postinjection. We finally employed
`the '1DMD mice to test human-specific AONs in vivo for
`their efficiency and specificity in inducing the skipping
`of exons 44, 46, and 49 from the human DMD gene.
`
`/
`
`I
`
`RESULTS
`Exon Skipping in Wild-Type Muscle
`We first set up targeted exon skipping in mouse muscle
`in vivo and optimized different parameters of adminis(cid:173)
`tration. We performed initial experiments in wild-type
`mice, and, while nonsense-mediated RNA decay will
`cause underestimation of the exon-skipping efficiencies,
`we monitored the effect of the AONs on the mRNA level
`only. We focused mainly on the specific skipping of
`exon 46, since in cultured mouse muscle cells an effec(cid:173)
`tive 2' -O-methylphosphorothioate oligoribonucleotide
`(m46AON4) was previously identified [17] . This AON
`was complexed to the delivery reagent PEI and injected
`into the gastrocnemius muscle of normal mice. Pilot
`experiments had indicated that highest skipping effi(cid:173)
`ciencies were reproducibly obtained following injections
`on 2 consecutive days (data not shown). We injected
`increasing dosages from 0.9 to 5.4 nmol m46AON4. RT(cid:173)
`PCR analysis of total muscle RNA demonstrated the
`occurrence of a novel shorter transcript fragment in all
`samples injected (Fig. IA). Sequence analysis confirmed
`the precise skipping of exon 46 in this product (data not
`shown). The skipping showed a slight dosage effect
`with plateau levels up to - 2.5% of total RT-PCR prod~
`ucts (in the entire muscle) following injection of 3.6
`nmol of the AON (Fig. lB). In subsequent experiments
`we observed an accumulation of skipping over time;
`analysis at 7 days postinjection revealed higher amounts
`of the skip product than after 2 days (data not shown).
`
`M OLECUL,\R THERAPY Vol. 10, No. 2, August 2004
`Copyright © The American Society of Gene Therapy
`
`233
`
`

`

`doi:10.1016/j.ymthe.2004.05 .031
`
`z
`
`0.9
`
`1.8
`
`3.6
`
`5.4
`
`w
`a.
`
`li:: I
`
`A.
`
`600 bp -
`l45 l46j47I ►
`145!471 ►
`
`B.
`
`3-----------------7
`ll 2.5
`0 g 2 + - - - - - - - - - - - - !
`:i e 1.5 +----------:=--------1
`!/I CL ~s
`B
`
`CL 'C
`
`0.5
`0 4--i------1-----1----l--~-1--+-+--,----1
`PEI
`5.4
`0.9
`NI
`3.6
`1.8
`Dose m46AON4 (nmol)
`
`FIG. 1. Dosage effects of antisense-induced exon 46
`skipping in muscle tissue of wild-type mice. (A) RT(cid:173)
`PCR analysis was performed at 7 days following
`injection of increasing doses of m46AON4 (0. 9 to
`5 .4 nmol, complexed to PEI) in both legs. A shorter
`transcript fragment (324 bp) was observed in all
`muscle samples injected. It was not present in muscle
`that was not injected (NI) or in muscle injected with
`PEI alone. The highest levels of exon skipping were
`detected at a dose of 3.6 nmol m46AON4. Due to
`heteroduplex formation we observed a slightly shorter
`fragment, migrating just below the wild-type product.
`M, 100 bp size marker; a negative RT-PCR control
`sample (-RT) was included. (B) The levels of exon 46
`skipping were determined through quantification of
`PCR products. The average ratios of skip products to
`total PCR products are indicated in percentages. At a
`dose of 3.6 nmol, the percentage of skip product was
`highest: 2.2%. A higher dose of 5.4 nmol appeared
`not to induce a higher skipping level.
`
`)
`
`In a time-series experiment we compared the effects of
`PEI and SAINT as delivery reagents to administrations of
`m46AON4 alone. We injected three series of mice with
`3.6 nmol of m46AON4, pure or complexed to PEI or
`SAINT, on 2 consecutive days. RT-PCR analysis was per(cid:173)
`formed at weekly time points, from day 1 to day 28
`postinjection. In all three series, we observed an increas(cid:173)
`ing level of skipped RNA product, reaching an optimum
`in the first to second week (Pig. 2A). PEI was clearly more
`efficient, showing highest skipping levels of up to 3% of
`total RT-PCR products. The skipping level subsequently
`diminished gradually, but was still clearly detectable after
`1 month. Compared to administration of the pure AON,
`the SAINT reagent appeared not to have a significantly
`enhancing effect.
`We analyzed cross sections of the contralateral
`injected muscles for dispersion and persistence of the
`fluorescein-labeled AON (Fig. 213). Following injection of
`the pure AON, we observed fluorescent signals within
`some fibers for up to 1 week. At later time points we
`observed only weak signals and mainly within the inter(cid:173)
`stitial spaces. The use of PEI clearly enhanced both
`dispersion and persistence of the fluorescent signal, even
`after 3 weeks. However, it also induced fiber degenera(cid:173)
`tion and monocyte infiltration absorbing most fluores(cid:173)
`cence. Using SAINT, most of the signal was detected in
`the interstitial spaces for up to 1 week, indicating that
`this reagent did not efficiently deliver the AON into the
`muscle fibers. Since the fluorescent signal may not
`correspond to the presence of intact and functional
`AONs, we performed MALDI-TOP mass spectrometry of
`
`injected muscle samples (Pig. 2C). The analyses indicated
`that the fluorescent label was removed from the AON
`within 24 h. The labeled AON was detectable only up to
`2 weeks when using PEI. The interstitial AONs were
`probably more vulnerable to degradation than the intra(cid:173)
`cellular AONs. The unlabeled AON was observed for 3 to
`4 weeks postinjection in all three series, but it may be
`functional only when present intracellularly, i.e., in the
`PEI series.
`
`Exon Skipping in Regenerating Muscle
`It is plausible to consider that the PEI-associated cyto(cid:173)
`toxic effect may actually have enhanced the delivery of
`the AONs. Hematoxylin/eosin analysis of cross sections
`of PEI-injected muscles showed significant fiber degen(cid:173)
`eration and monocyte infiltration in the first week
`postinjection, followed by extensive regeneration, as
`determined by the number of centrally nucleated fibers
`(Fig. 3). At day 28, a significant fraction of the muscle
`was still regenerating. Using antibodies against CD4 and
`CD8, we identified the infiltrating cells as cytotoxic and
`helper T cells (data not shown). We observed no signif(cid:173)
`icant cytotoxic response following injections of pure
`m46AON4 or m46AON4 complexed with SAINT (data
`not shown). The enhancing effect of fiber regeneration
`on AON delivery suggests that the myofiber membrane
`is a significant barrier. Therefore, we compared exon(cid:173)
`skipping efficiencies at decreasing closes of m46AON4 in
`normal muscle with those in mdx muscle, which exhib(cid:173)
`its extensive fiber de- and regeneration. Indeed, at an
`equal dose of 3.6 nmol, we observed almost 3-fold
`
`234
`
`MOLECUL,\R Ti IERAPY Vol. 10, No. 2, August 2004
`Copyright © The American Society of Gene Therapy
`
`

`

`doi:10.1016/j.ymthe.2004.05 .031
`
`A.
`
`ID AON D AON/PEI • AON/SAINT I
`
`(/J
`
`3,5
`.... u
`3
`::, 2,5
`0
`a. "O
`...
`2
`·- 0
`.II:
`(/J a. 1,5
`~s
`1 .
`....
`0,5
`0
`
`0
`
`B.
`
`AON
`
`AON
`+ PEI
`
`AON
`+SAINT
`
`C.
`
`1600
`
`1200
`
`800
`
`400
`
`0
`
`1600
`
`1200
`
`800
`
`400
`
`d1
`
`d1
`
`d21
`d14
`d7
`Time points post-injection
`
`d28
`
`d7
`
`d14
`
`d21
`
`t-<
`
`0
`
`N "'
`
`l()
`
`4000 5000 6000
`Control muscle
`
`0
`0
`It)
`
`"'
`
`0
`N
`
`"'
`
`It)
`
`0 ,-..
`
`l()
`l()
`
`4000 5000 6000
`6-FAM-AON
`
`0
`It)
`
`"'
`"'
`
`0 00
`,-.. N
`0
`l ( ) " '
`l() It)
`l()
`
`"'
`
`4000
`
`3000
`
`2000
`
`1000
`
`0
`
`1200
`
`800
`
`400
`
`3000
`
`2000
`
`1000
`
`0
`
`1600
`
`1200
`
`800
`
`400
`
`0
`
`"' 0
`
`It)
`
`4000 5000 6000
`AON
`
`0
`0
`It)
`
`"'
`
`0
`N
`
`"'
`
`l()
`
`4000 5000 6000
`d1AON
`
`0
`
`6000
`5000
`4000
`d1 AON+PEI
`
`0
`3000 4000 5000 6000
`d1 AON+SAINT
`
`6-FAM-AON 5570
`
`AON
`
`5030
`
`X
`
`X X X
`
`5570
`
`5030
`
`X X
`X X X X X
`
`5570
`
`5030
`
`X
`X X X
`
`AONtr
`
`3950
`
`X
`
`3950
`
`X
`
`X
`
`3950
`
`X X
`
`d1
`
`d7 d14 d21 d28
`AON
`
`d1 d7 d14 d21 d28
`AON+ PEI
`
`d1 d7 d14 d21 d28
`AON+ SAINT
`
`FIG. 2. Time-response analysis in wild-type
`mice following intramuscular injections of 3.6
`nmol m46AON4, either pure or complexed
`to PEI or SAINT. (A) RT-PCR analyses were
`performed at different time points postin(cid:173)
`jection (days 1, 7, 14, 21, and 28). The levels
`of exon 46 skipping were determined
`through quantification of PCR products. The
`average ratios of skip products to total PCR
`products are indicated as percentages. In all
`three series the exon 46 skipping levels
`increased over time, to highest levels in the
`first to second week postinjection. PEI was
`most efficient in delivering the AON, as
`indicated by the skipping levels that were
`highest (2.8%) at 14 days and most persist(cid:173)
`ing for up to 28 days. Injections of pure
`m46AON4 or complexed to the SAINT
`reagent were markedly less efficient com(cid:173)
`pared to those with PEI. (B) lmmunohisto(cid:173)
`chemical analyses of cross sections from the
`injected gastrocnemius muscles at increasing
`time points postinjection (up to 21 days). The
`desmin staining (red fluorescence) indicates
`the fiber structure. The AON is detected by its
`green fluorescent label. Following injection of
`pure m46AON4, some positive fibers were
`observed for up to 7 days. Using PEI as
`delivery reagent, a markedly higher level of
`fluorescence was observed both within the
`muscle fibers and in the interstitial spaces for
`at least 21 days. The infiltrating mono(cid:173)
`nucleated cells also absorbed significant
`fluorescence. SAINT did not significantly
`enhance the uptake of m46AON4, and most
`fluorescence was detected in between the
`fibers, for up to 7 days. Original magnifica(cid:173)
`tion, 200 x ; scale bar, 30 r,m. (C) MALDI-TOF
`mass spectrometry of muscle samples from
`the time series to assess the intramuscular
`persistence of m46AON4 after delivery. The
`molecular weights of unlabeled versus la(cid:173)
`beled m46AON4 are 5030 and 5570 daltons
`respectively (top). In all samples a non'.
`specific, unidentified, fragment of MW 5920
`was sometimes observed, which was also
`present in the control (not injected) muscle
`sample. The middle shows individual analy(cid:173)
`ses of muscle samples at day 1 following
`injection of m46AON4, either pure or com(cid:173)
`plexed to PEI or SAINT. In contrast to the
`other samples in which the fluorescent label
`was removed within 24 h, the labeled
`m46AON4 was detected in the PEI series
`only. The bottom represents diagrammatic
`overviews of the three different time-series
`experiments, showing the long-term persis(cid:173)
`tence of m46AON4, which was best (up to
`28 days) following PEI delivery. A degrada(cid:173)
`tion product (AONtr) of MW 3950 was
`sometimes seen. Its composition and residual
`function are unspecified.
`
`MOLECULAR TIJER,IJ>Y Vol. 10, No. 2, August 2004
`Copyright <0 The American Society of Gene Therapy
`
`235
`
`

`

`doi:10.1016/j.ymthe.2004.05.031
`
`FIG. 3. Histological analysis (hematoxylin/eosin stain(cid:173)
`ing) of cross sections from the gastrocnemius muscles
`injected with m46AON4 complexed to the PEI delivery
`reagent in the time-series experiment (see Fig. 2). At
`days 1, 4, and 7 postinjection, massive cell infiltration
`and muscle fiber degeneration were observed. At later
`time points (up to 28 days), extensive muscle
`regeneration occurred, as indicated by the number of
`centrally nuclea ted fibers. Original magnification,
`200 x; scale bar, 30 µm.
`
`higher levels of exon 46 skipping (up to 9%) in mdx
`muscle (Figs. 4A and 4B). Furthermore, we determined
`the minimal effective dose to be 10 times lower.
`
`Human-Specific Exon Skipping in hDMD Muscle
`Since the exon-skipping strategy is a sequence-specific
`therapeutic approach, the ideal preclinical validation
`would be a target human DMD gene, in a mouse exper(cid:173)
`imental background. We have engineered such transge(cid:173)
`nic, "humanized" DMD (hDMD) mice carrying an
`integrated and functional copy of the full-length human
`
`DMD gene. We detected expression of human dystrophin
`in hDMD mouse muscle specifically by immunohisto(cid:173)
`chemical analysis of cross sections (Fig. SA), using a
`human-specific antibody (MANDYS106). On the muscle
`RNA level, RT-PCR analyses using either mouse- or hu(cid:173)
`man-specific primers demonstrated correct transcription
`of the human DMD gene (Fig. SB and unpublished
`results). Furthermore, upon crossing with mdx mice, the
`hDMD construct was shown to complement the dystro(cid:173)
`phic defect, as was assessed by histological a?d cDNA
`microarray analysis (manuscript in preparation). We
`
`FIG . 4. Antisense-induced exon 46 skipping in
`regenerating muscle of mdx mice. (A) RT-PCR analysis
`was performed at 7 days postinjection of increasing
`doses of m46AON4 (0.36 to 3.6 nmol, complexed to
`PEI) in both legs. Exon 46 skipping, represented by a
`novel shorter transcript fragment (324 bp), was
`observed in all mdx muscle samples injected. It was
`not present in mdx muscle that was not injected (NI).
`Compared to 3.6 nmol m46AON4 injected in muscle
`of wild-type (control) mice, the skipping levels in mdx
`mice were significantly higher. M, 100 bp size marker;
`a negative RT-PCR control sample (-RT) was included.
`(B) The levels of exon 46 skipping were determined
`through quantification of PCR products. The average
`ratios of skip products to total PCR products are
`indicated in percentages. At a dose of 3.6 nmol, the
`percentage of skip product was almost 3-fold higher in
`mdx versus normal mice. At the lowest dose of 0.36
`nmol, exon 46 skipping was still detectable at levels of
`approximately 1%.
`
`236
`
`mdx
`
`control
`
`0.36
`
`0.9
`
`1.8
`
`3.6
`
`3.6
`
`z
`
`0 mdx
`II control
`
`A.
`
`600 bp -
`1451461471 ►
`145(47! ►
`
`B.
`
`12
`
`10
`.l!l
`... CJ
`::I 8
`0
`C. 'C
`0
`-
`6
`.lO: ..
`II) C.
`~g 4
`
`0 .. 2
`
`0
`
`NI
`
`0.36
`
`1.8
`3.6
`0.9
`Dose m46AON4 (nmol)
`
`3.6
`
`Th is mate,ri.a I was co,.pied
`at the NLM a,nd may be
`Subject US Copyright Laws
`
`MOLECULAR TIIER,\PY Vol. 10, No. 2, August 2004
`Copyright © The American Society of Gene Therapy
`
`

`

`doi:10.1016/". mthe.2004.05.031
`
`Mouse
`
`hDMD
`
`Human
`
`A.
`
`MANDYS
`106
`
`Dys2
`
`B.
`
`FIG. 5. The transgenic hDMD m ice as a model
`to study human-specific e xon skipping. (A)
`lmmunohistochemical analysis of cross sec(cid:173)
`tions from human, wild-type mouse, and
`hDMD mouse muscle. Using the h uman(cid:173)
`specific antibody MANDYSl 06, human dystro(cid:173)
`ph in was detected in hu man and hDMD mo use
`muscle, w he reas no protein was detected in
`wild-type mouse muscle (to p). This indicated
`that the integrated copy of the human DMD
`gene is correctly expressed in the transge nic
`hDMD mice. Using DYS2, which cross-reacts
`with both h uman and mouse dystrophin,
`membrane signals w ere detected in all sections
`(bottom). Original magnificatio n, 200x; scale
`bar, 30 rim. (B) RT-PCR a nalysis o f RNA samples
`from hDMD muscles 10 days after injection of
`the human-specific h44AON1 , h46AON8, or
`h49AON 1, versus control human (H), mouse
`(M), and hDMD mouse samples (al l not
`injected). The human-specific PCRs (left) show
`the specific AON-induced skipping of the
`ap p ropriate target exons from the human
`DMD transcripts, whereas the mouse-specific
`PCRs (rig