sign in sign up
<<
Home , Adenovirus genome, Viral vector

Gene Therapy (2005) 12, S18–S27 & 2005 Nature Publishing Group All rights reserved 0969-7128/05 $30.00 www.nature.com/gt CONFERENCE PAPER Gutless adenovirus: last-generation adenovirus for therapy

R Alba1, A Bosch1 and M Chillon1,2 1Gene Therapy Laboratory, Department of Biochemistry and , Center of Biotechnology and (CBATEG), Universitat Auto`noma de Barcelona, Bellaterra, Spain; and 2Institut Catala` de Recerca i Estudis Avanc¸ats (ICREA), Barcelona, Spain

Last-generation adenovirus vectors, also called helper-depen- viral coding regions, gutless vectors require viral dent or gutless adenovirus, are very attractive for gene therapy supplied in trans by a helper . To remove contamination because the associated in vivo is highly by a from the final preparation, different systems reduced compared to first- and second-generation adenovirus based on the excision of the helper-packaging signal have vectors, while maintaining high efficiency and been generated. Among them, Cre-loxP system is mostly tropism. Nowadays, gutless adenovirus is administered in used, although contamination levels still are 0.1–1% too high different organs, such as the liver, muscle or the central to be used in clinical trials. Recently developed strategies to nervous system achieving high-level and long-term transgene avoid/reduce helper contamination were reviewed. expression in rodents and primates. However, as devoid of all Gene Therapy (2005) 12, S18–S27. doi:10.1038/sj.gt.3302612

Keywords: adenovirus; gutless; helper-dependent vectors; in vivo gene therapy

Introduction clinical for more information). Nowadays, adenovirus vectors are applied to treat , monogenic disorders, Gene therapy for most genetic diseases requires expres- vascular diseases and others complications. sion of the therapeutic for the whole of Basically, wild-type adenovirus is associated with the patient. In order to be efficient for the treatment mild diseases like conjunctivitis, pharyngitis and, in a of genetic disorders, a gene therapy vector has to meet high percentage, with colds or acute respiratory diseases several conditions: (i) safety, which can be better but not to tumoral pathways or other viral alterations.1 In achieved with a nonintegrative vector, as it avoids the the early 1950s, adenovirus was used for the first time risk for ; (ii) ability to be easily as a preventive for respiratory diseases. These and inexpensively produced at a large-scale in the studies supposed a great advance since they allowed the laboratory; (iii) stability in target cells, which is favored knowledge of aspects related to associated immune with low-immunogenic vectors; and (iv) high-capacity responses and adverse secondary effects. allowing the possibility of introducing full-length DNA As gene transfer vector, the adenovirus has a high sequences for most , long full-length cDNAs, efficiency, both in quiescent and in dividing endogenous promoters or additional regulatory cells, it does not integrate; allows easy modification sequences such as enhancers or insulators, which can in order to retarget its tropism to different tissues; and provide a tightly regulated expression of the therapeutic high titers (up to 1013 particles/ml) are routinely obtained. gene, similar to physiologic conditions. Adenoviruses In addition, it is noteworthy that in the last 10 years, new and especially gutless adenoviruses seem to accomplish scale-up adenovirus production systems have been most of these conditions as they are not integrative, they developed, facilitating its use in human clinical trials. have a capacity of 36 kb, they can be easily produced at Well-characterized human serotypes Ad2 and Ad5 high titers in the laboratory and can be delivered to a from group C are the classic adenoviruses used as large mass of cells, and, contrary to the first-generation vectors. In order to produce safe and nonreplicative adenovirus in which the expression of wild-type adeno- vectors, first-generation adenoviral vectors containing viral genes stimulates the immune system, gutless the whole viral with the exception of the E1 adenoviruses show long-term stability in many tissues. region were developed.2 To propagate first-generation As seen in Table 1, adenovirus is one of the first elections adenoviruses, several E1-expressing lines have been in clinical trials, being the most used vector in the last generated: 293,3 911,4 N52.E65 and PER.C6.6 5 years (1999–2004) (visit www.wiley.co.uk/genmed/ Although E1-deleted vectors cannot replicate in vivo, residual expression from adenoviral genes triggers a cytotoxic T lymphocyte (CTL) immune response towards Correspondence: Dr M Chillo´n, Gene Therapy Laboratory, Department of 7 Biochemistry and Molecular Biology, Center of Animal Biotechnology and infected cells, which finally leads to the elimination of Gene Therapy (CBATEG), Edifici H, Universitat Auto`noma de Barcelona, transduced cells and, therefore, to the lost of therapeutic Bellaterra 08193, Spain . Gutless adenovirus R Alba et al S19 To avoid this problem, second-generation adenoviral non-human DNA.14 However, first injections of gutless vectors combining deletion of different early regions vectors carrying lambda DNA elicited CTL immune (E17E3 and E2/E4) were generated (Figure 1). Deletion response because peptides from stuffer were presented of these regions permits to accommodate up to 14 kb and in the cell membrane, and it became evident that DNA hence increase the vector cloning capacity.8,9 However, stuffer played an important role in the stabilization of second-generation Ad vectors still do not avoid in vivo- viral DNA into the cell.15,16 The first choice was then associated immunogenicity and toxicity due to residual DNA from mammalian and human introns since they gene expression from remaining viral genes. seemed to favor maintenance of the gutless genome into Third-generation vectors, called gutless or gutted the cell for long periods of time.15 For example, intronic Ad, devoid of all coding viral regions were recently sequences from the HPRT gene containing matrix generated. They are also called helper-dependent adeno- attachment regions (MAR) elements have been used as because of the need of a helper adenovirus that stuffer DNA showing increased DNA stability and no carries all coding regions, and high-capacity adenoviruses induction of a CTL response. Sequences from other because they can accommodate up to 36 kb of DNA. human loci have also been used with similar or better Briefly, the gutless adenovirus only keeps the 50 and 30 results.17 inverted terminal repeats (ITRs) and the packaging However, not all intronic sequences are appropriate signal (C) from the wild-type adenovirus. DNA stuffer and, thus, proper candidates should avoid Vector package efficiently only 75–105% of the the following sequences (follow next characteristics) whole adenovirus genome.10–14 As therapeutic expres- (i) coding regions, (ii) repetitive sequences (like alu sion cassettes usually do not reach up to 36 kb, there is a sequences), (iii) hot spots for recombination, (iv) regions need to use stuffer DNA in order to complete the genome that can interfere with the expression of the transgene, size for encapsidation. Initially, it was believed that and (v) toxic or immunogenic regions. On the contrary, stuffer DNA’s unique role was the participation in the MAR that seem to stabilize the adenovirus genome into packaging of the vector genome. Thus, the first DNA the nucleus and permit long periods of gene expression stuffer used was from , yeast, bacterial and are recommended.17

Table 1 Vectors used in gene therapy clinical trials Production of gutless adenovirus vectors Total number of Protocol number Since gutless vectors are devoid of all viral genes, protocols (2004) increase (1999–2004) proteins needed for its genome replication, packaging and capsid formation must be supplied in trans. This 263 83 is achieved by coinfection of the gutless with a helper Adenovirus 258 172 Lipofection 85 10 adenovirus. However, since both helper and gutless Adeno-associated virus 25 21 vectors have the same viral capsid, separation must be Other viruses 148 100 addressed before purification. Thus far, strategies have Others 183 148 been based on reducing the packaging efficiency of the helper genome compared to the gutless genome, either Data from www.wiley.co.uk/genmed/clinical. by mutating its packaging signal,18–20 by the different

0 10000 20000 30000 36000

MLP Ψ E1L1 L2 L3 L4 E3 L5 Ad5 genome ITR ITR E2B E2A E4

Ψ ∆E1 ∆E3 First generation Transgene

Ψ ∆E1 ∆E2 ∆E3 ∆E4 Second generation Transgene

Ψ Helper Dependent Ad Transgene Figure 1 Map of adenovirus serotype 5 genome and different generations of adenoviral vectors. Early transcripts are represented by E1–E4 regions and late transcripts are represented by L1–L5 regions. MLP: major late promoter; C: packaging signal.

Gene Therapy Gutless adenovirus R Alba et al S20 size of its genome13 ( bigger or smaller than These levels of helper contamination seem to be due the optimal do not package efficiently), or by specific to the limited efficiency of packaging signal excision elimination of its packaging signal during viral produc- associated to low recombinase activity or with low tion.21 endogenous levels of recombinases. This is caused either Initial strategies consisted in the use of helper- by adenovirus-mediated host cell shut off or by dependent adenoviruses carrying a defective packaging cytotoxicity of high levels of Cre recombinase.22,23 signal.13 These vectors had extensive regions of deleted Although the final level of helper contamination is low, viral genome, but they still have some coding regions. further reduction is desirable to minimize any potential Cotransfection of a wild-type adenovirus as a helper toxicity associated with the helper virus, especially when with this gutless permitted to propagate both adeno- high doses are required.24 viruses. After several passages, the gutless adenovirus Other recombinases such as FLP have also been used

was purified by CsCl2 density gradient. However, high to remove the packaging signal. FLP is a yeast site- levels of helper virus and multiple recombinations specific recombinase, which catalyzes recombination between both vectors were detected. This, together with between frt sites.25 Philip Ng and collaborators26 created important complications in large-scale production and new FLP-expressing cell lines (293FLP and 293CreFLP) purification, led to the development of new strategies. and a helper adenovirus with a packaging signal flanked An important advance was the specific removal of the by frt sites. Cre and FLP seem to have equivalent packaging signal by Cre recombinase. Parks and colla- efficiency in the production of gutless adenovirus with a borators21 developed a production system where the final helper contamination of 0.1–1%. A recent and helper adenovirus had a packaging signal flanked by two elegant improvement in the Cre-loxP system has been loxP sites, and amplification was performed in Cre the reversion of the packaging signal of the helper virus recombinase-expressing cell lines (293Cre). When the in order to avoid generation of a replication competent helper adenovirus entered the cell, its packaging signal adenovirus (RCA) by recombination with the gutless was excised preventing the inclusion of its genome into vector, which has reduced the helper contamination to the viral particle, but retaining all coding regions for the levels down to 0.02–0.1%.18 However, Cre and FLP are viral proteins needed to produce the gutless vectors bidirectional recombinases that permit excision of the (Figure 2). Using this system, average helper contamina- packaging signal and also its re-entrance, favoring tion ranged over 0.1–10% compared to gutless vectors. contamination by the helper adenovirus. Improving

HELPER Ad GUTLESS Ad GENOME ITRΨ ITR ITR ITR ADENOVIRUS GENOME Ψ HUMAN STUFFER DNA

loxP Sequences

PLASMID OR VIRUS COTRANSFECTION

Ψ + Ψ +

RECOMBINASE Ψ + CRE

VIRAL Ψ - PROTEINS

CRE/293 CELL LINE

O.1-10%

HELPER ADENOVIRUS GUTLESS ADENOVIRUS CONTAMINATION

Figure 2 Generation of gutless adenovirus using the Cre/loxP system. Gutless and helper genomes are cotransfected in permissive 293 Cre-expressing cells, where both genomes are amplified and viral proteins produced. Then, packaging signal of the helper’s genome is excised by Cre recombinase, preventing its packaging into the viral capsid, while gutless genome is still packageable. Efficiency of the excision process allows 90–99.9% purity of the gutless vector.

Gene Therapy Gutless adenovirus R Alba et al S21 the excision efficiency of Cre and FLP in cell lines cells and the generation of neutralizing antibodies also improves the opposite reaction. Thus, the use of against the transgene products and the adenovirus unidirectional recombinases, such as FC31, is an capsid. Both nonspecific innate and adaptive immune attractive alternative to reduce contamination levels with responses are involved when first- and second-genera- the helper adenovirus. FC31 is a unidirectional recombi- tion adenoviral vectors are administered (see Schagen nase, which recognizes homologous sequences called et al33 for an extensive review). Thus, the innate immune attB/attP and shows an excision efficiency similar to that response is rapidly developed after virus entry by of Cre in 293 cells.27 When FC31 excises attB/attP- induction of inflammatory gene expression and further flanked sequences, it creates new sequences called attR/ recruitment of macrophages, neutrophil and natural attL preventing the opposite reaction. Thus, the helper killer cells, leading to an 80–90% of first-generation adenovirus carrying an attB/attP-flanked packaging vector removal from the liver in 24 h.34 Basically, innate signal opens the door to a new gutless production is triggered by the adenovirus particle, is system (authors’ unpublished results). Ad-dose dependent and does not require viral gene Another approach to reduce helper Ad contamination expression.35–37 is based on size-restricted protein IX-deleted helper. pIX In a second step, adaptive cellular and humoral is a hexon-associated protein essential for packaging full- immune responses are developed about 4–7 days after length viral genomes. A 293 pIX-expressing cell line delivery. At this time, a second peak of cytokine and permitting the growth of a full-length pIX-defective chemokine gene expression and inflammation occurs helper was generated.28 The use of gutless vectors with leading to lymphocytic infiltrates and to the induction genomes smaller than 36 kb gives a selective advantage of adenovirus-specific CTL.35 Initially, cellular immune in the pIX system. Plaque-forming unit assay showed response is activated when -presenting cells a reduction by 1000 times of helper contamination (APCs) uptake adenovirus particles, process the particles compared to the classical Cre-loxP system. However, into small oligopeptides and present them through the southern blot analysis revealed helper contamination major histocompatibility complex (MHC) class-I mole- levels of 0.2% (similar to Cre-loxP), meaning that the cules at the cell surface. Further binding of CD8+ T cells majority of helper adenoviruses was packaged into the to the MHC class-I/peptide complex induces formation capsid. Combining both Cre and pIX size-restricted of Ad-or transgene-product-specific CTLs. Therefore, the systems allowed to reduce the number of helper novo synthesis does not seem to be required to initiate the adenovirus particles, but defective helper virions still process.33 However, for late inflammation, the expression remain in the final preparation. of viral genes still encoded within Ad vectors plays To address the helper contamination issue, other a significant role. In immunocompetent hosts, this strategies are based on using nonadenovirus vectors as response limits the duration of transgene expression helpers, like herpes simplex virus-129 and baculovirus.30 and results in adenovirus vector clearance within a few However, low production efficiency29 and 2% of RCA weeks of administration.7,34 generation30 make their use not feasible in large-scale On the other hand, adaptive humoral immune production. response is initiated by the binding of adenovirus As important as the design of a helper-free gutless particles to the surface immunoglobulin of B cells.33 adenovirus system is its production in large or industrial After internalization and virus processing, the adeno- scale. Large-scale production of gutless adenovirus is virus-derived epitopes are presented at the surface of the complex and less efficient since it requires successive B cell by MHC-II molecules. Exposure of these cells to time-consuming gutless:helper adenovirus coinfections. cytokines from activated CD4+-Th2 helper cells will Unfortunately, this process increases the risk of reorga- result in differentiated plasma cells secreting antibodies nizations in the vector genome by homologous recombi- towards the adenoviral capsid.38 High titers of antibodies nation with viral E1 sequences present in permissive 293 against capsid proteins, either pre-existing because of cells, which can finally lead to the generation of RCAs. To previous exposure to natural virus or generated as a avoid this, several laboratories have developed permis- result of vector administration, may inhibit subsequent sive cells, such as N52.E65 and PERC6,6 that do not dosing with the same vector. contain viral sequences prone to recombine. However, to Different strategies to circumvent innate and adaptive achieve large quantities of high quality helper-free immune responses have been developed. However, most gutless vectors needed for clinical assays, it is funda- of them present secondary complications and/or their mental to develop and improve the methodology in the use in human patients is questionable. These strategies amplification and purification steps as well as in vector include macrophage depletion,39,40 use of immunosup- quality. Thus, development of several cell lines able to pressive agents (cyclosporin A, cyclophosphamide, grow in suspension, like human 293S,31 PER.C6,32 and dexamethasone, FK506, Interleukin-12 and deoxyper- 293Cre suspension cell lines18 facilitates large-scale gualin),41–46 use of antibodies to deplete CTLs,47,48 amplification in bioreactors, although purification meth- blockade of costimulatory interactions between APCs, ods like non-ultracentrifuge-based methods and elimina- T and B cells,49–52 intrathymic administration of adeno- tion of helper contamination still need to be improved. virus,53 oral tolerization,54 use of vectors derived from non-crossreacting serotypes,55,56 use of adenoviruses from other species57,58 and coating vectors with inert 59–61 Gutless adenovirus and immune response chemicals like polyethylene glycol (PEG). Diverse in vivo studies in mice suggested that, in the Systemic delivery of first-generation adenoviral vectors absence of an immune response, first-generation adeno- is known to induce a strong host’s immune response, DNA is maintained as a stable episome in the resulting in the rapid elimination of vector-transduced host cell.41,62,63 Last-generation helper-dependent or

Gene Therapy Gutless adenovirus R Alba et al S22 gutless adenovirus vectors display reduced long-term Surprisingly, gutless vectors have also been proved to toxicity and prolonged transgene expression compared be very efficient in , due to their longer to first-generation vectors after administration to per- duration of expression, their lower antiviral reactivity ipheral organs of immunologically naı¨ve .56,64–68 and their higher levels of transgene protein in dendritic Lack of coding viral genes may account for reduced cells compared to the same amount of first-generation adaptive cellular immune response after systemic deliv- Ad vectors.79 ery of gutless vectors. Initially, gutless vectors are capable of transducing dendritic cells and stimulating Ad-specific T-cell responses, independent of viral gene In vivo administration of gutless adenovirus .69 However, the expression of viral genes is required for T cells to exert their effector functions in the Gutless adenoviruses have been administered in vivo to liver,70 which possibly explains the vector persistence different tissues in rodents, dogs and also to non-human and improved transgene expression following transduc- primates. Owing to the high efficiency of adenoviruses in tion with gutless vectors compared to results with first- targeting hepatocytes, most of the toxicity studies were generation Ad vectors. carried out in the liver, following intravenous adminis- As expected, systemic delivery of gutless vectors still tration of the vector.68 Gutless vectors have been shown induces adaptive humoral response against the vector to express non-immunogenic transgenes during the capsid as for first-generation Ad vectors. Indeed, the whole life of the mouse, while expression driven development of Ad-specific antibodies does not con- by first-generation adenovirus lasted at most for 3 tribute to the elimination of Ad-transduced cells and months.80–82 therefore does not affect the persistence of transgene Therapeutic genes carried by gutless vectors have expression. However, Ad-specific antibodies will bind been administered to the liver of mouse models for the readministered Ad vector and thereby prevent cell different diseases, such as hemophilia A and B,83,84 entry and promote opsonization by macrophages. obesity,66 familial hypercholesterolemia,81,85–87 ornithine Humoral response can also be developed towards transcarbamylase deficiency,88 diabetes89 and chronic circulating antigen induced by gutless adenoviral trans- viral hepatitis.90,91 Therapeutic levels of most proteins fer.70 Vectors that mediate transgene expression in APCs have been documented for a long-term duration. In some trigger antibody formation because they increase the cases the antibody response against a nonendogenous probability of neoantigen presentation by APCs,71 and, transgene led to a significant decrease in the circulating hence, careful selection of tissue-specific promoters may levels of the therapeutic protein. However, when the significantly improve adenovirus-associated toxicity transgene expression was driven by a liver-specific profiles and diminish or abolish APC transduction and promoter, the immune response was lower, efficacy of transgene expression.72 In addition, systemic adminis- the therapeutic protein increased and secondary effects tration of gutless vectors in a clinical setting might be derived from systemic circulation of the therapeutic inefficient because of the presence of circulating neu- protein were undetectable.90,92 The encouraging results tralizing antibodies against the same or crossreactive obtained in rodents had promoted the preclinical studies serotype as a consequence of a natural infection or as a in larger animal models. This is the case for hemophilia result of previous vector administration. A large propor- A and B dogs,72,83 which resulted in minimal acute liver tion of the human population has significant levels of toxicity and of the transgene antibodies against adenoviruses;73 thus, the capacity to allowing for partial or complete correction of hemo- overcome a pre-existing immunity is a fundamental philia. As in rodents, the use of liver-specific promoters problem. Different successful strategies to circumvent avoided the development of neutralizing antibodies pre-existing immunity have been applied, such as the against the therapeutic protein, although the expression use of alternative gutless serotypes15 and the use of a was lower than with a viral ubiquitous promoter.72 non-human gutless adenovirus.58 Thus, administration Long-term expression of gutless-driven transgene of gutless CAV-2 vectors allows the stable, high-level was also documented in baboons for more than 1 year; expression in throughout the rat central nervous however, a decrease in the levels of the protein was system (CNS).20 It is noteworthy that even in the observed over time.56 Toxicity in non-human primates presence of an active peripheral immunization with an was assessed using different doses of gutless adenovirus adenovirus that completely eliminates expression from and the results obtained showed that administration of first-generation vectors within 60 days, gutless vectors high doses of adenovirus (1013 viral particles/kg) are able to maintain a long-term transgene expression induced a strong activation of the innate inflammatory in the brain74,75 due to reduced inflammation both in response that caused the death of the animal.76 intensity and in duration in the brains of the preimmu- In order to circumvent the decrease in transgene nized animals. expression overtime, readministration with gutless vec- As innate immune responses are dependent on the tors has been attempted in mice using different adeno- viral capsid or particle, innate responses stimulated by virus serotypes. While readministration with different gutless vectors are similar to those stimulated by first- serotypes is possible using first-generation adeno- generation adenovirus vectors. Thus, dose-dependent virus,55,56,93–95 the level of transgene expression is lower acute inflammation was reported by Brunetti-Pierri and after the second administration, probably due to cross- colleagues76 in non-human primates following the reacting CTLs between the different serotype proteins.96 administration of high-dose gutless vectors. However, However, as gutless vectors do not contain viral genes, innate response, as these recently reported, may be readministration of a different serotype leads to the same reduced by PEG modification, probably due to lower levels of the therapeutic protein as in naı¨ve animals vector uptake by Kupffer cells in vivo.77,78 without showing liver toxicity.15,81 Liver toxicity was also

Gene Therapy Gutless adenovirus R Alba et al S23 absent after repeated administration of the same ser- that the decline in DNA copies could be due to instability otype of gutless;67 however, it resulted in a drop of 30- of the vector DNA into the host cell.104 However, loss of to 100-fold in transgene expression compared to naı¨ve DNA copy number/cell has no consequences in dystro- animals.15 phin expression or muscle function in experiments Efficiency of regulated promoters was also evaluated performed in mice because dystrophin is stable for 26 using gutless vectors as they offer the possibility of weeks. When skeletal muscle fibers were transduced incorporating all the genetic components of an inducible before birth (in utero at embryonic day 16), expression of system plus other additional sequences such as insula- the reporter protein was stable at least for 5 months tors or silencers that have been shown to avoid leakiness despite the development of antibodies against the of the uninduced promoter.97,98 When these promoters transgene and the adenoviral capsid.106 However, in this were driven by first-generation adenoviruses, only one case, the authors reported no loss of DNA copy number. induction was possible; a significant hepatotoxicity was Lower levels of toxicity compared to first-generation developed when the animals were treated with the eneration Ad that resulted in higher survival rates of inducer, followed by a rapid loss of vector DNA.99 This animals transduced in utero were also documented. was not the case with gutless vectors, where expression Mdx muscle goes through cycles of necrosis followed was reinduced three and four times over a period of 2 by regeneration, which could account for the higher months. Similar results were obtained with different efficiency of infection of muscle fibers from young and inducible systems such as tetracycline, tamoxifen or old mdx mice by adenovirus vectors compared to control mifepristone.97–99 animals as reported by the group of JS Chamberlain.107 Stability of gutless vectors was evaluated during On the contrary, the high turnover of muscle cells in the hepatocyte cell cycle. Surprisingly, after two-third partial mdx mice can decrease the stability of nonintegrative hepatectomy, gutless vector genomes were reduced by vectors like adenovirus. In 1-year old mdx mice, 50%, while first-generation Ad vector copies were transduction of 25–30% of skeletal muscle cross-sectional reduced by 71%. More importantly, episomal area with gutless vectors carrying the dystrophin cDNA DNA-injected mice showed a reduction of 99% in DNA under the regulation of a muscle-specific promoter lead copy number. The decrease in vector copy number/cell to 40% correction of their high susceptibility of contrac- correlated with transgene expression. Several hypotheses tion-induced injury 1 month after treatment.107 However, have been proposed to explain the persistence of longer time points were not evaluated, but mild immune episomal gutless vectors in dividing hepatocytes, like cell infiltration was reported using mouse dystrophin centromeric function of the Ad genome or nuclear cDNA. retention activity of the Ad terminal protein, but Administration of recombinant protein to the CNS conclusive results have not been obtained to date.100 requires the breakdown of the blood–brain barrier, a Adenoviruses target the liver very efficiently; how- procedure that should not be used for repetitive read- ever, they are less effective in other cell types lacking the mnistrations. Thus, permanent transduction of neurons CAR or adenovirus receptor. Retargeting of gutless offers the possibility of stable treatment of diseases vectors has been attempted by the incorporation of affecting the CNS. Gene transfer to the CNS has been polylysine or RGD in the H–I loop of the adenoviral fiber explored using different vector systems. Despite the protein, which can be obtained by adding fiber-modified immunoprotection of the CNS, results with first-genera- helper viruses in the last amplification step of the gutless tion adenovirus show a decrease in the expression of the vector production. Polylysine has been shown to transgene 2 months after transduction, correlating with effectively transduce mature muscle cells and the disappearance of the adenoviral DNA.108 Acute loss in vivo,101 while RGD increased the efficiency of of transduced cells and chronic inflammation could transducing ovarian carcinoma cell lines, primary account for this decline, mainly at high doses.109 On the vascular smooth muscle cells and primary human contrary, gutless adenovirus expression can be detected endothelial cells.102 in the CNS 1 year after striatal injections.20,110 Moreover Duchenne muscular dystrophy (DMD), an X-linked and contrary to what was reported for other tissues, lethal disorder that affects 1 in 3500 males, is caused by preimmunization with the same serotype of adenovirus genetic in the dystrophin gene. The cDNA for does not alter gutless stability in the CNS, and does not full-length dystrophin is approximately of 14 kb, far significantly reduce the initial infection of the CNS by above the size of most gene transfer viral vectors. Thus, gutless vectors, although a transient acute brain inflam- the highcapacity of gutless adenoviruses has opened the mation is shown.75 Gutless vectors have also been used possibility to treat DMD animal models103 not only with to successfully express therapeutic genes in the retina one full-length dystrophin cDNA but also with two either by transplanting cells previously transduced with copies of the therapeutic gene.104,105 In these studies, the virus or by direct injection of the vector.110,111 neonate skeletal muscles of mdx mice injected with Other tissues like lung, cardiac muscle, vascular tissue gutless adenovirus expressed dystrophin for the dura- or dendritic cells have also been targeted with gutless tion of the experiment, up to 1 year. At this point, 52% adenoviruses with similar results as for the muscle or the of the muscle fibers showed transgene expression and liver.112–115 functional correction of muscle contractility, and improved histopathology was reported, correlating with the level of dystrophin expression. However, a four-fold Future considerations decline in the vector DNA copy number was observed from 10 days to 1 year after vector injection, and the This is a fascinating moment for gutless adenovirus treated mice developed a significant humoral response. vectors: their use permits long-term expression in vivo No transduced muscle fiber loss was reported, indicating with reduced and transient cellular immune response in

Gene Therapy Gutless adenovirus R Alba et al S24 animal models for human diseases and, moreover, the 14 Parks RJ, Graham FL. A helper-dependent system for adeno- increasing number of groups working in this field favors virus vector production helps define a lower limit for efficient technological advances to escape preimmune response DNA packaging. J Virol 1997; 71: 3293–3298. by developing non-human gutless adenovirus20 or to 15 Parks R, Evelegh C, Graham F. Use of helper-dependent increase stability by generating integrative gutless adenoviral vectors of alternative serotypes permits repeat vectors,116,117 or vectors with replication capacity.118 vector administration. Gene Therapy 1999; 6: 1565–1573. However, their use in clinical assays is questionable 16 Schiedner G et al. Variables affecting in vivo performance of since helper contamination levels are still high, and high-capacity adenovirus vectors. J Virol 2002; 76: 1600–1609. large-scale production in bioreactors is not yet fully 17 Sandig V et al. Optimization of the helper-dependent adeno- developed. Therefore, future studies will be needed in virus system for production and potency in vivo. Proc Natl Acad 97 these areas to finally bring gutless adenovirus to the Sci USA 2000; : 1002–1007. 18 Palmer D, Ng P. Improved system for helper-dependent clinic. adenoviral vector production. Mol Ther 2003; 8: 846–852. 19 Sato M, Suzuki S, Kubo S, Mitani K. Replication and packaging of helper-dependent adenoviral vectors. Gene Therapy 2002; 9: Acknowledgements 472–476. 20 Soudais C, Skander N, Kremer EJ. Long-term in vivo transduc- We would like to acknowledge Dr Merce` Monfar for tion of neurons throughout the rat CNS using novel helper- critically reading the manuscript. Our work is supported dependent CAV-2 vectors. FASEB J 2004; 18: 391–393. by MCYT-SAF2003-03256, Marato´ TV3-2002-031632 and 21 Parks RJ et al. A helper-dependent adenovirus vector system: Instituto de Salud Carlos III (C03/08). AB has a contract removal of helper virus by Cre-mediated excision of the viral from the Ramon y Cajal Program (Ministerio Educacio´n packaging signal. Proc Natl Acad Sci USA 1996; 93: 13565–13570. y Ciencia, Spain), and RA is a recipient of an FI- 22 Silver DP, Livingston DM. Self-excising retroviral vectors Generalitat fellowship. encoding the Cre recombinase overcome Cre-mediated cellular toxicity. Mol Cell 2001; 8: 233–243. 23 Zang Y, Schneider R. Adenovirus inhibition of cellular protein synthesis and the specific of late viral mRNA. Semin References Virol 1993; 4: 233–243. 24 Ng PG, Graham FL. Helper-dependent adenoviral vectors for 1 Horwitz MS. and their replication. 1990; gene therapy. In: Templeton NS (ed). Gene and Cell Therapy. 2: 1679–1720. Marcel Dekker Inc: New York, 2004 pp 53–70. 2 Danthinne X, Imperiale MJ. Production of first generation 25 Som T, Armstrong KA, Volkert FC, Broach JR. Autoregulation adenovirus vectors: a review. Gene Therapy 2000; 7: 1707–1714. of 2 micron circle gene expression provides a model for 3 Graham FL, Smiley J, Russell WC, Nairn R. Characteristics of a maintenance of stable plasmid copy levels. Cell 1988; 52: 27–37. human cell line transformed by DNA from human adenovirus 26 Ng P et al. Development of a FLP/frt system for generating type 5. J Gen Virol 1977; 36: 59–74. helper-dependent adenoviral vectors. Mol Ther 2001; 3: 809–815. 4 Fallaux FJ et al. Characterization of 911: a new helper cell line 27 Groth AC, Olivares EC, Thyagarajan B, Calos MP. A phage for the titration and propagation of early region 1-deleted directs efficient site-specific integration in human adenoviral vectors. Hum Gene Ther 1996; 7: 215–222. cells. Proc Natl Acad Sci USA 2000; 97: 5995–6000. 5 Schiedner G, Hertel S, Kochanek S. Efficient transformation of 28 Sargent KL et al. Development of a size-restricted pIX-deleted primary human amniocytes by E1 functions of Ad5: generation helper virus for amplification of helper-dependent adenovirus of new cell lines for adenoviral vector production. Hum Gene vectors. Gene Therapy 2004; 11: 504–511. Ther 2000; 11: 2105–2116. 29 Kubo S, Saeki Y, Chiocca EA, Mitani K. An HSV amplicon- 6 Fallaux FJ et al. New helper cells and matched early region based helper system for helper-dependent adenoviral vectors. 1-deleted adenovirus vectors prevent generation of replication- Biochem Biophys Res Commun 2003; 307: 826–830. competent adenoviruses. Hum Gene Ther 1998; 9: 1909–1917. 30 Cheshenko N, Krougliak N, Eisensmith RC, Krougliak VA. A 7 Yang Y et al. Cellular immunity to viral limits novel system for the production of fully deleted adenovirus E1-deleted adenoviruses for gene therapy. Proc Natl Acad Sci vectors that does not require helper adenovirus. Gene Therapy USA 1994; 91: 4407–4411. 2001; 8: 846–854. 8 Amalfitano A et al. Production and characterization of 31 Graham FL. Growth of 293 cells in suspension culture. J Gen improved adenovirus vectors with the E1, E2b, and E3 genes Virol 1987; 68 (Pt 3): 937–940. deleted. J Virol 1998; 72: 926–933. 32 Sakhuja K et al. Optimization of the generation and propagation 9 Armentano D et al. Effect of the E4 region on the persistence of of gutless adenoviral vectors. Hum Gene Ther 2003; 14: 243–254. transgene expression from adenovirus vectors. J Virol 1997; 71: 33 Schagen FH, Ossevoort M, Toes RE, Hoeben RC. Immune 2408–2416. responses against adenoviral vectors and their transgene 10 Alemany R et al. Complementation of helper-dependent products: a review of strategies for evasion. Crit Rev Oncol adenoviral vectors: size effects and titer fluctuations. J Virol Hematol 2004; 50: 51–70. Methods 1997; 68: 147–159. 34 Worgall S, Wolff G, Falck-Pedersen E, Crystal RG. Innate immune 11 Bett AJ, Prevec L, Graham FL. Packaging capacity and stability mechanisms dominate elimination of adenoviral vectors follow- of human adenovirus type 5 vectors. J Virol 1993; 67: 5911–5921. ing in vivo administration. Hum Gene Ther 1997; 8: 37–44. 12 Ghosh-Choudhury G, Graham FL. Stable transfer of a mouse 35 Liu Q et al. The role of capsid-endothelial interactions in the dihydrofolate reductase gene into a deficient cell line using innate immune response to adenovirus vectors. Hum Gene Ther human adenovirus vector. Biochem Biophys Res Commun 1987; 2003; 14: 627–643. 147: 964–973. 36 Schnell MA et al. Activation of innate immunity in nonhuman 13 Mitani K, Graham FL, Caskey CT, Kochanek S. Rescue, primates following intraportal administration of adenoviral propagation, and partial purification of a helper virus- vectors. Mol Ther 2001; 3: 708–722. dependent adenovirus vector. Proc Natl Acad Sci USA 1995; 37 Molinier-Frenkel V et al. Immune response to recombinant 92: 3854–3858. adenovirus in humans: capsid components from viral input are

Gene Therapy Gutless adenovirus R Alba et al S25 targets for vector-specific cytotoxic T lymphocytes. J Virol 2000; 56 Morral N et al. Administration of helper-dependent adenoviral 74: 7678–7682. vectors and sequential delivery of different vector serotype for 38 Paul WE, Seder RA. Lymphocyte responses and cytokines. Cell long-term liver-directed gene transfer in baboons. Proc Natl 1994; 76: 241–251. Acad Sci USA 1999; 96: 12816–12821. 39 Kuzmin AI, Finegold MJ, Eisensmith RC. Macrophage deple- 57 Fitzgerald JC et al. A simian replication-defective adenoviral tion increases the safety, efficacy and persistence of adenovirus- recombinant vaccine to HIV-1 gag. J Immunol 2003; 170: mediated gene transfer in vivo. Gene Therapy 1997; 4: 309–316. 1416–1422. 40 Wolff G et al. Enhancement of in vivo adenovirus-mediated gene 58 Kremer EJ, Boutin S, Chillon M, Danos O. Canine adenovirus transfer and expression by prior depletion of tissue macro- vectors: an alternative for adenovirus-mediated gene transfer. phages in the target organ. J Virol 1997; 71: 624–629. J Virol 2000; 74: 505–512. 41 Dai Y et al. Cellular and humoral immune responses to 59 Chillon M, Lee JH, Fasbender A, Welsh MJ. Adenovirus adenoviral vectors containing factor IX gene: tolerization of complexed with polyethylene glycol and cationic lipid is factor IX and vector antigens allows for long-term expression. shielded from neutralizing antibodies in vitro. Gene Therapy Proc Natl Acad Sci USA 1995; 92: 1401–1405. 1998; 5: 995–1002. 42 Fang B et al. Gene therapy for hemophilia B: host immunosup- 60 Croyle MA, Chirmule N, Zhang Y, Wilson JM. ‘Stealth’ pression prolongs the therapeutic effect of adenovirus-mediated adenoviruses blunt cell-mediated and humoral immune re- factor IX expression. Hum Gene Ther 1995; 6: 1039–1044. sponses against the virus and allow for significant gene 43 Kaplan JM, Smith AE. Transient immunosuppression with expression upon readministration in the lung. J Virol 2001; 75: deoxyspergualin improves longevity of transgene expression 4792–4801. and ability to readminister adenoviral vector to the mouse 61 Croyle MA, Chirmule N, Zhang Y, Wilson JM. PEGylation lung. Hum Gene Ther 1997; 8: 1095–1104. of E1-deleted adenovirus vectors allows significant gene 44 Kuriyama S et al. Immunomodulation with FK506 around the expression on readministration to liver. Hum Gene Ther 2002; time of intravenous re-administration of an adenoviral vector 13: 1887–1900. facilitates gene transfer into primed rat liver. Int J Cancer 2000; 62 Lochmuller H et al. Immunosuppression by FK506 markedly 85: 839–844. prolongs expression of adenovirus-delivered transgene in 45 Otake K, Ennist DL, Harrod K, Trapnell BC. Nonspecific skeletal muscles of adult dystrophic [mdx] mice. Biochem inflammation inhibits adenovirus-mediated pulmonary gene Biophys Res Commun 1995; 213: 569–574. transfer and expression independent of specific acquired 63 Vilquin JT et al. FK506 immunosuppression to control the immune responses. Hum Gene Ther 1998; 9: 2207–2222. immune reactions triggered by first-generation adenovirus- 46 Zuckerman JB et al. A phase I study of adenovirus-mediated mediated gene transfer. Hum Gene Ther 1995; 6: 1391–1401. transfer of the human cystic fibrosis transmembrane conduc- 64 Chen HH et al. DNA from both high-capacity and first- tance regulator gene to a lung segment of individuals with generation adenoviral vectors remains intact in skeletal muscle. cystic fibrosis. Hum Gene Ther 1999; 10: 2973–2985. Hum Gene Ther 1999; 10: 365–373. 47 Poller W et al. Stabilization of transgene expression by 65 Maione D et al. Prolonged expression and effective read- incorporation of E3 region genes into an adenoviral factor IX ministration of erythropoietin delivered with a fully deleted vector and by transient anti-CD4 treatment of the host. Gene adenoviral vector. Hum Gene Ther 2000; 11: 859–868. Therapy 1996; 3: 521–530. 66 Morsy MA et al. An adenoviral vector deleted for all viral 48 Sawchuk SJ et al. Anti-T cell receptor monoclonal antibody coding sequences results in enhanced safety and extended prolongs transgene expression following adenovirus-mediated expression of a leptin transgene. Proc Natl Acad Sci USA 1998; in vivo gene transfer to mouse synovium. Hum Gene Ther 1996; 95: 7866–7871. 7: 499–506. 67 O’Neal WK et al. Toxicity associated with repeated adminis- 49 Kay MA et al. Long-term hepatic adenovirus-mediated gene tration of first-generation adenovirus vectors does not occur expression in mice following CTLA4Ig administration. Nat with a helper-dependent vector. Mol Med 2000; 6: 179–195. Genet 1995; 11: 191–197. 68 Schiedner G et al. Genomic DNA transfer with a high-capacity 50 Kay MA et al. Transient immunomodulation with anti-CD40 adenovirus vector results in improved in vivo gene expression ligand antibody and CTLA4Ig enhances persistence and and decreased toxicity. Nat Genet 1998; 18: 180–183. secondary adenovirus-mediated gene transfer into mouse liver. 69 Roth MD et al. Helper-dependent adenoviral vectors efficiently Proc Natl Acad Sci USA 1997; 94: 4686–4691. express transgenes in human dendritic cells but still stimulate 51 Stein CS, Pemberton JL, van Rooijen N, Davidson BL. Effects of antiviral immune responses. J Immunol 2002; 169: 4651–4656. macrophage depletion and anti-CD40 ligand on transgene 70 Muruve DA et al. Helper-dependent adenovirus vectors elicit expression and redosing with recombinant adenovirus. Gene intact innate but attenuated adaptive host immune responses in Therapy 1998; 5: 431–439. vivo. J Virol 2004; 78: 5966–5972. 52 Wilson CB et al. Transient inhibition of CD28 and CD40 ligand 71 De Geest BR, Van Linthout SA, Collen D. Humoral immune interactions prolongs adenovirus-mediated transgene expres- response in mice against a circulating antigen induced by sion in the lung and facilitates expression after secondary adenoviral transfer is strictly dependent on expression in vector administration. J Virol 1998; 72: 7542–7550. antigen-presenting cells. Blood 2003; 101: 2551–2556. 53 DeMatteo RP et al. Long-lasting adenovirus transgene expression 72 Brown BD et al. Factors influencing therapeutic efficacy and the in mice through neonatal intrathymic tolerance induction with- host immune response to helper-dependent adenoviral gene out the use of immunosuppression. JVirol1997; 71: 5330–5335. therapy in hemophilia A mice. J Thromb Haemost 2004; 2: 54 Ilan Y et al. Oral tolerization to adenoviral proteins permits 111–118. repeated adenovirus-mediated gene therapy in rats with 73 Kremer EJ. CAR chasing: canine adenovirus vectors-all bite and pre-existing immunity to adenoviruses. Hepatology 1998; 27: no bark? J Gene Med 2004; 6 (Suppl 1): S139–S151. 1368–1376. 74 Thomas CE et al. Peripheral infection with adenovirus causes 55 Mastrangeli A et al. ‘Sero-switch’ adenovirus-mediated in vivo unexpected long-term brain inflammation in animals injected gene transfer: circumvention of anti-adenovirus humoral intracranially with first-generation, but not with high-capacity, immune defenses against repeat adenovirus vector adminis- adenovirus vectors: toward realistic long-term neurological tration by changing the adenovirus serotype. Hum Gene Ther gene therapy for chronic diseases. Proc Natl Acad Sci USA 2000; 1996; 7: 79–87. 97: 7482–7487.

Gene Therapy Gutless adenovirus R Alba et al S26 75 Thomas CE et al. Pre-existing antiadenoviral immunity is not a 95 Roy S, Shirley PS, McClelland A, Kaleko M. Circumvention of barrier to efficient and stable transduction of the brain, immunity to the adenovirus major coat protein hexon. J Virol mediated by novel high-capacity adenovirus vectors. Hum 1998; 72: 6875–6879. Gene Ther 2001; 12: 839–846. 96 Smith CA, Woodruff LS, Rooney C, Kitchingman GR. Extensive 76 Brunetti-Pierri N et al. Acute toxicity after high-dose systemic cross-reactivity of adenovirus-specific cytotoxic T cells. Hum injection of helper-dependent adenoviral vectors into non- Gene Ther 1998; 9: 1419–1427. human primates. Hum Gene Ther 2004; 15: 35–46. 97 Burcin MM et al. Adenovirus-mediated regulable target gene 77 Croyle MA et al. PEGylated helper-dependent adenoviral expression in vivo. Proc Natl Acad Sci USA 1999; 96: 355–360. vectors: highly efficient vectors with an enhanced safety profile. 98 Salucci V et al. Tight control of gene expression by a helper- Gene Therapy 2005; 12: 579–587. dependent adenovirus vector carrying the rtTA2(s)-M2 tetra- 78 Mok H, Palmer DJ, Ng P, Barry MA. Evaluation of polyethylene cycline transactivator and repressor system. Gene Therapy 2002; glycol modification of first-generation and helper-dependent 9: 1415–1421. adenoviral vectors to reduce innate immune responses. Mol 99 Zerby D et al. In vivo ligand-inducible regulation of gene Ther 2005; 11: 66–79. expression in a gutless adenoviral vector system. Hum Gene 79 Harui A et al. Vaccination with helper-dependent adenovirus Ther 2003; 14: 749–761. enhances the generation of transgene-specific CTL. Gene 100 Ehrhardt A, Xu H, Kay MA. Episomal persistence of Therapy 2004; 11: 1617–1626. recombinant adenoviral vector genomes during the cell cycle 80 Chen HH et al. Persistence in muscle of an adenoviral vector in vivo. J Virol 2003; 77: 7689–7695. that lacks all viral genes. Proc Natl Acad Sci USA 1997; 94: 101 Bramson JL et al. Helper-dependent adenoviral vectors contain- 1645–1650. ing modified fiber for improved transduction of developing 81 Kim IA et al. Potential of adenoviral p53 gene therapy and and mature muscle cells. Hum Gene Ther 2004; 15: 179–188. irradiation for the treatment of malignant gliomas. Int J Oncol 102 Biermann V et al. Targeting of high-capacity adenoviral vectors. 2001; 19: 1041–1047. Hum Gene Ther 2001; 12: 1757–1769. 82 Morral N et al. High doses of a helper-dependent adenoviral 103 Bulfield G, Siller WG, Wight PA, Moore KJ. X - vector yield supraphysiological levels of alpha1-antitrypsin linked muscular dystrophy (mdx) in the mouse. Proc Natl Acad with negligible toxicity. Hum Gene Ther 1998; 9: 2709–2716. Sci USA 1984; 81: 1189–1192. 83 Ehrhardt A et al. A gene-deleted adenoviral vector results 104 Dudley RW et al. Sustained improvement of muscle function in phenotypic correction of canine hemophilia B without one year after full-length dystrophin gene transfer into mdx liver toxicity or thrombocytopenia. Blood 2003; 102: mice by a gutted helper-dependent adenoviral vector. Hum 2403–2411. Gene Ther 2004; 15: 145–156. 84 Reddy PS et al. Sustained human factor VIII expression in 105 Gilbert R et al. Prolonged dystrophin expression and functional hemophilia A mice following systemic delivery of a gutless correction of mdx mouse muscle following gene transfer with a adenoviral vector. Mol Ther 2002; 5: 63–73. helper-dependent (gutted) adenovirus-encoding murine dys- 85 Belalcazar LM et al. Long-term stable expression of human trophin. Hum Mol Genet 2003; 12: 1287–1299. apolipoprotein A–I mediated by helper-dependent adenovirus 106 Bilbao R et al. Comparison of high-capacity and first-generation gene transfer inhibits atherosclerosis progression and remodels adenoviral vector to murine muscle in utero. atherosclerotic plaques in a mouse model of familial hyper- Gene Therapy 2005; 12: 39–47. cholesterolemia. Circulation 2003; 107: 2726–2732. 107 DelloRusso C et al. Functional correction of adult mdx mouse 86 Oka K et al. Long-term stable correction of low-density muscle using gutted adenoviral vectors expressing full-length lipoprotein receptor-deficient mice with a helper-dependent dystrophin. Proc Natl Acad Sci USA 2002; 99: 12979–12984. adenoviral vector expressing the very low-density lipoprotein 108 Zou L, Zhou H, Pastore L, Yang K. Prolonged transgene receptor. Circulation 2001; 103: 1274–1281. expression mediated by a helper-dependent adenoviral vector 87 Pastore L et al. Helper-dependent adenoviral vector-mediated (hdAd) in the central nervous system. Mol Ther 2000; 2: 105–113. long-term expression of human apolipoprotein A–I reduces 109 Thomas CE et al. Acute direct adenoviral vector cytotoxicity atherosclerosis in apo E-deficient mice. Gene 2004; 327: 153–160. and chronic, but not acute, inflammatory responses correlate 88 Mian A et al. Long-term correction of ornithine transcarbamy- with decreased vector-mediated transgene expression in the lase deficiency by WPRE-mediated overexpression using a brain. Mol Ther 2001; 3: 36–46. helper-dependent adenovirus. Mol Ther 2004; 10: 492–499. 110 Semkova I et al. Autologous transplantation of genetically 89 Kojima H et al. NeuroD-betacellulin gene therapy induces islet modified iris pigment epithelial cells: a promising concept neogenesis in the liver and reverses diabetes in mice. Nat Med for the treatment of age-related macular degeneration and 2003; 9: 596–603. other disorders of the eye. Proc Natl Acad Sci USA 2002; 99: 90 Aurisicchio L et al. Liver-specific alpha 2 interferon gene 13090–13095. expression results in protection from induced hepatitis. J Virol 111 Oshima Y et al. Intraocular gutless adenoviral-vectored VEGF 2000; 74: 4816–4823. stimulates anterior segment but not retinal neovascularization. 91 Fiedler M et al. Helper-dependent adenoviral vector-mediated J Cell Physiol 2004; 199: 399–411. delivery of woodchuck-specific genes for alpha interferon (IFN- 112 Fleury S et al. Helper-dependent adenovirus vectors devoid of alpha) and IFN-gamma: IFN-alpha but not IFN-gamma reduces all viral genes cause less myocardial inflammation compared woodchuck hepatitis virus replication in chronic infection in with first-generation adenovirus vectors. Basic Res Cardiol 2004; vivo. J Virol 2004; 78: 10111–10121. 99: 247–256. 92 Pastore L et al. Use of a liver-specific promoter reduces immune 113 Tuettenberg A et al. Early adenoviral gene expression mediates response to the transgene in adenoviral vectors. Hum Gene Ther immunosuppression by transduced dendritic cell (DC): im- 1999; 10: 1773–1781. plications for using genetically modified DC. 93 Kass-Eisler A et al. Circumventing the immune response J Immunol 2004; 172: 1524–1530. to adenovirus-mediated gene therapy. Gene Therapy 1996; 3: 114 Wen S, Graf S, Massey PG, Dichek DA. Improved vascular gene 154–162. transfer with a helper-dependent adenoviral vector. Circulation 94 Mack CA et al. Circumvention of anti-adenovirus neutralizing 2004; 110: 1484–1491. immunity by administration of an adenoviral vector of an 115 Koehler DR et al. Protection of Cftr knockout mice from acute alternate serotype. Hum Gene Ther 1997; 8: 99–109. lung infection by a helper-dependent adenoviral vector

Gene Therapy Gutless adenovirus R Alba et al S27 expressing Cftr in airway epithelia. Proc Natl Acad Sci USA 117 Yant SR et al. Transposition from a gutless adeno-transposon 2003; 100: 15364–15369. vector stabilizes transgene expression in vivo. Nat Biotechnol 116 Goncalves MA et al. Transfer of the full-length dystrophin- 2002; 20: 999–1005. coding sequence into muscle cells by a dual high-capacity 118 Kreppel F, Kochanek S. Long-term transgene expression in hybrid viral vector with site-specific integration ability. J Virol proliferating cells mediated by episomally maintained high- 2005; 79: 3146–3162. capacity adenovirus vectors. J Virol 2004; 78: 9–22.

Gene Therapy