Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Gene Therapy (1999) 6, 637–642 1999 Stockton Press All rights reserved 0969-7128/99 $12.00 http://www.stockton-press.co.uk/gt Inhibition of adenovirus-mediated gene transfer by bronchoalveolar lavage fluid A Bastian and B Bewig First Department of Internal Medicine, Christian-Albrechts-University of Kiel, Schittenhelmstrae 12, 24105 Kiel, Germany Human and animal trials with recombinant adenovirus have been discouraging, since the level of recombinant gene expression was low. Nonspecific and specific immune response mediated by, for example, macrophages, T cells and immunoglobulins may prevent infection or cause death of infected cells. We analyzed the effect of bronchoalveolar lavage fluid (BAL) on the efficiency of adenoviral infection in vitro. A total of 26 BAL samples of randomly selected patients was examined. Adenovirus-mediated gene transfer was quantified using AdCMV.Null, a recombinant adenovirus, in a modified titer assay based on immunocytochemical detection of infected 293 cells. In addition, the concentration of anti-adenovirustype 5 IgA, IgG and IgM antibodies in BAL was determined by ELISA. 53.8% of the BAL samples (14 out of 26) reduced adenoviral infectivity by at least 50% (factor of inhibition ⭓2). All BAL samples effecting a reduction in adenoviral infectivity contained detectable amounts of antiadenovirus-type 5-IgA antibodies. However, the correlation between the concentration of IgA antibody and the strength of inhibition was weak (r = 0.336). Even high levels of antiadenovirus-type 5-IgM, IgG or IgA antibodies did not influence adenoviral infectivity consistently. This observation indicates that BAL contains (a) anti-adenovirus-type 5 antibodies which are not directed against adenoviral epitopes responsible for the viral adherence and uptake process; and/or (b) inhibitors of viral infectivity different from antibodies. Keywords: BAL; adenovirus; gene transfer; immunoglobulin Introduction Gene therapy may become a powerful tool for the treatment of inherited or acquired genetic diseases.1–4 A key problem is the generation of efficient gene transfer vectors and strategies. Success of adenovirus-mediated gene transfer is limited by the development of host immune response resulting in diminished or lacking gene expression. Two phases of elimination have been described. Introduction of recombinant adenovirus into mammalian airways stimulates destructive cellular responses, leading to the replacement of the airway with non-transgene-containing cells. This process is thought to be responsible for the late phase of elimination.5–7 Modification of the adenovirus with inactivation of the E2a gene in addition to E1 sequences was associated with longer gene expression and less inflammation in vivo, indicating possible strategies of circumvention.8,9 However, the recently observed early elimination of virus has major implications, since 90% of vector DNA is eliminated within 24 h following in vivo administration.10 Macrophages are known to be involved in this process. In addition, antibody-mediated immune response, primed during a prior wild-type adenovirus infection, is likely to reduce the initial efficiency of gene transfer. Depending on the route of administration, special local features may have implications on the efficiency of adenoviral gene transfer. For the treatment of pulmonary manifestations Correspondence: B Bewig Received 24 March 1998; accepted 4 November 1998 of genetic diseases, application via the airways seems promising, since this route provides direct and specific access to epithelial cells of the bronchi. However, adenovirus administered via the airways has to cross the epithelial lining fluid barrier before reaching the target cell. No qualitative or quantitative data are available with regard to how adenovirus infection is influenced by noncellular substances in the bronchoalveolar fluid (BAL). We examined the effect of BAL from unselected patients on the infectivity of recombinant adenovirus in vitro using a modified neutralization assay. Results from these experiments were analyzed in relation to a quantitative assessment of specific immunoglobulin A, M and G titers in BAL directed specifically against recombinant adenovirus type 5, which is frequently used in gene transfer experiments. Results Effect of BAL fluid on adenoviral infection of cells in vitro To evaluate the effect of BAL fluid on the infectivity of recombinant adenovirus in vitro, a modified neutralization test was used. 293 Cells were incubated with normal saline as control or with BAL samples obtained from different patients. The cells were then infected with recombinant adenovirus at a dose of 0.001 p.f.u. per cell. Infected cells were detected immunocytochemically and counted. The number of infected cells after incubation with BAL samples was compared to controls (Figure 1). Of 26 BAL samples, 14 demonstrated significant inhi- Inhibition of adenovirus-mediated gene transfer by BAL A Bastian and B Bewig 638 Table 1 Properties of samples from bronchoalveolar lavage fluid Factor of inhibition 22.5 18.0 13.7 9.3 8.4 7.4 7.3 7.3 6.5 4.5 3.4 Figure 1 Effect of bronchoalveolar lavage samples on the infectivity of recombinant adenovirus in vitro. 293 Cells were grown as monolayer in 96-well plates. After incubation with bronchoalveolar lavage sample (20 g protein) from unselected patients or normal saline as control, cells were infected with a recombinant adenovirus AdCMV.Null at a multiplicity of infection of 0.001 p.f.u. per cell. Infected cells were stained immunocytochemically utilizing an antibody specifically directed against adenoviral proteins 48 h after infection and counted using light microscopy. Points indicate percentage change of number of infected cells for each individual patient when incubated with BAL as compared with cells incubated with control solution. 2.4 2.3 2.0 1.9 1.9 1.8 1.6 1.5 1.2 1.0 1.0 1.0 1.0 0.9 0.8 Disease sarcoidosis healthy sarcoidosis sarcoidosis sarcoidosis tuberculosis tuberculosis carconoma sarcoidosis healthy lymphocytic pneumonitis sarcoidosis lymphoma healthy sarcoidosis carcinoma carcinoma bronchitis sarcoidosis carcinoma healthy lymphoma sarcoidosis asthma asthma healthy IgA-AK quant (ng/90 g protein) Ad-IgG-AK Ad-IgM(extinction) Ak (extinction) 36.0 16.0 33.0 91.0 22.0 91.0 8.5 68.0 87.0 54.0 23.5 0.31 0.55 0.82 1.32 0.68 1.00 0.64 0.84 1.47 1.25 0.49 0.03 0.17 0.19 0.15 0.01 0.25 0.10 0.19 0.12 0.24 0.09 49.0 9.2 21.0 12.8 28.5 10.7 0.0 0.0 18.0 21.0 7.7 7.1 7.1 20.0 4.4 0.68 0.77 1.17 0.71 1.02 1.88 0.82 0.15 0.90 0.25 0.21 1.25 0.52 1.37 0.24 0.00 0.06 1.13 0.00 0.00 0.16 0.07 0.02 0.04 0.00 0.02 0.57 0.00 0.50 0.00 bition of adenoviral infection in 293 cells. The extent of inhibition varied from factor 2 to 22.5, demonstrating a reduction of infected cells to at least 50% in the presence of BAL compared with noninfected cells. BAL samples from three different patients each reduced the number of infected cells to less than 10% compared with controls. In 10 samples, BAL was found to have no or little influence on infectivity of adenovirus (less than 50% inhibition). Two BAL specimens were observed to mediate a slight increase in adenovirus uptake into cells. While three patients who suffered from sarcoidosis had BAL samples with no significant effect on adenoviral infectivity, four out of the five samples demonstrating the most powerful inhibition of adenoviral transfer were obtained from patients with sarcoidosis (Table 1). to the inhibitory properties of the BAL specimens (Figure 2). Most samples (20 out of 26) contained detectable amounts of antibodies against adenovirus type 5. BAL samples with no effect on adenoviral infectivity contained varying titers of IgA antibodies (between 0 and 87 ng/20 g protein). All samples but one inhibiting adenovirus infection to less than 20% compared with control (factor of inhibition less than 5) contained less than 30 ng IgA/20 g BAL protein, one of the samples had 49 ng IgA/20 g protein. There seems to be a correlation between inhibitory effect and concentration of IgA up to a factor of inhibition of 9.3 and IgA of 91 ng/20 g protein. However, the most powerful inhibitory effects, which clearly exceeded these samples, were obtained with samples of medium range IgA titers as high as 36 ng IgA/20 g protein. One sample demonstrating significant inhibition of infectivity contained only low IgA anti adenovirus antibodies of 8.5 ng IgA/20 g protein, a range where other samples displayed no inhibitory effect. Overall correlation was r = 0.336. Inhibition of adenoviral infectivity and detection of IgA antibodies against adenovirus in BAL All samples examined in the infectivity assay were analyzed for the presence of antibodies specifically directed against recombinant adenovirus type 5, the same adenovirus used in the infectivity assay. IgA antibody content in BAL was determined using ELISA technique. Results were expressed as immunoglobulin content relative to the protein concentration of the BAL samples and related Inhibition of adenoviral infectivity and detection of IgG antibodies against adenovirus in BAL In order to evaluate the relation between IgG specific for adenovirus type 5 and the effect of BAL on adenoviral infection of cells, the specific IgG level in BAL was determined by ELISA and results were correlated with the results from the infectivity assay. There was a wide range of anti-adenovirus IgG antibody concentrations in BAL samples having no effect on the extent of adenoviral Inhibition of adenovirus-mediated gene transfer by BAL A Bastian and B Bewig 639 Figure 2 Concentration of IgA-specific against adenovirus type 5 in relation to the factor of inhibition of adenoviral infection in concentrated BAL. BAL from unselected patients was rendered cell-free by centrifugation. Protein content was adjusted to 1 g/l. IgA concentration was determined using ELISA technique in a sandwich with an adenovirus type 5 covered titer plate, BAL sample and anti-human IgA-specific antibody and a standard curve using IgA. Factor of inhibition was evaluated in an infectivity assay using 293-cells incubated with BAL sample (20 g protein) or normal saline as control. Cells were infected with AdCMV.Null at a multiplicity of infection of 0.001 p.f.u. per cell. Infected cells were stained immunocytochemically utilizing an antibody specifically directed against adenoviral proteins 48 h after infection and using light microscopy. The factor of inhibition was defined as the ratio of the number of infected cells in the control to the number of infected cells in the BAL sample. Each symbol represents the mean of two values. Figure 3 Concentration of IgG specific against adenovirus type 5 in relation to the factor of inhibition of adenoviral infection in concentrated BAL. Analysis was performed as described for Figure 2, but IgG concentration was determined using ELISA technique in a sandwich with an adenovirus type 5 covered titer plate, BAL sample and anti-human IgG specific antibody. Data are presented as extinction read at 405 nm in the ELISA assay compared with the factor of inhibition. Each symbol represents the mean of two values. infection (Figure 3). The samples with the highest and the lowest IgG titer had both similar effects on adenoviral infectivity. Mid-range inhibition of infectivity was associated with mid-range IgG titers. The BAL sample which inhibited infection most strongly, contained low IgG antibody titers comparable to others with no inhibiting potential. Inhibition of adenoviral infectivity and detection of IgM antibodies against adenovirus in BAL Evaluation of BAL samples for the presence of specific anti-adenovirus type 5-IgM antibody titer by ELISA technique revealed low level, but detectable amounts of antibodies in most of the patients. Results were expressed in relation to the extent of inhibition the corresponding BAL sample exhibited in the infectivity assay. The overall level of IgM directed against adenovirus type 5 in BAL was low (Figure 4). There were only three samples demonstrating increased IgM levels clearly different from the rest. Analysis showed no correlation between concentration of specific anti-adenovirus type 5-IgM and the influence of the BAL sample on the infectivity of the adenovirus in vitro. Even the patient with the highest antibody titer against adenovirus type 5 did not display relevant change in the infectivity assay if cells were pre- Figure 4 Concentration of IgM specific against adenovirus type 5 in relation to the factor of inhibition of adenoviral infection in concentrated BAL. Analysis was performed as described for Figure 2, but IgM concentration was determined using ELISA technique in a sandwich with an adenovirus type 5 covered titer plate, BAL sample and anti-human IgM specific antibody. Results are presented as extinction read at 405 nm in the ELISA assay compared with the factor of inhibition. Each symbol represents the mean of two values. Inhibition of adenovirus-mediated gene transfer by BAL A Bastian and B Bewig 640 incubated with patient’s BAL. The patient with a BAL which most strongly inhibited adenoviral infection, had low antibody titers in the range of other patients with no inhibition. BAL samples from all three patients with significant anti-adenovirus IgM antibodies had no potential to inhibit adenoviral infectivity. Discussion Numerous applications revealed adenovirus as a powerful gene transfer vector in vitro. However, success of in vivo gene therapy utilizing adenovirus as the vector was limited by the transient expression of the transferred gene. Several studies analyzing adenovirus-based gene transfer to different organs in animals and humans demonstrated progressive loss of recombinant gene expression over days and weeks.2–5,11–14 The reason for this phenomenon was in part the development of an adaptive immune response with generation of antigenspecific cytotoxic T cells recognizing processed viral or recombinant gene products on the surface of infected cells. These cytotoxic T cells are thought to destroy infected cells directly or to mediate apoptosis.5,6,14–17 Consequently, expression from the adenovirus vector was markedly prolonged in nude mice lacking the T cellbased immune response.17 Strategies to circumvent T cellmediated host defense included modified viral vectors with diminished expression of viral antigens or enhanced expression of genes suppressing MHC presentation on infected cells.18 Even immunosuppressive agents (eg FK506) were used to prevent cellular immune response.19 Recent studies suggest a more rapid elimination of adenovirus DNA from the target organ within the first 24 h.10 After systemic vector administration to mice, the vast majority of vector genome was cleared in the liver. Innate actions of the immune system like unspecific barrier functions, humoral or cell-mediated functions were thought to be responsible. However, the precise location (eg Kupffer cells) and mechanism of vector elimination remains to be determined. Depending on the route of administration, local effects possibly inhibiting expression from adenoviral vectors are of certain interest. Gene transfer into the lung, a possible target organ for gene therapy in cystic fibrosis, can be achieved via the airways. In this scenario systemic clearance is of no relevance. However, adenovirus may face a variety of cellular and non-cellular barriers within the epithelial lining fluid before finally entering the epithelial cells. Alveolar macrophages as part of the innate immune system and T-lymphocytes may be involved in eliminating intruding viruses. Our study analyzed the role of the non-cellular fraction of the BAL in modulating the infectivity of recombinant adenovirus. The data demonstrate that some BAL samples contain components reducing the infectivity of recombinant adenovirus. Most BAL samples with the property of reducing adenoviral infectivity contained detectable amounts of anti-adenovirus-type 5-IgA antibodies. However, concentration of IgA antibody did not correlate with the strength of inhibition. Animal and human studies indicate the development of humoral immune response after wild-type or recombinant adenoviral infection. In a mouse model, airway infection with adenovirus type 5 produced neutralizing antibodies in sera and BAL. Isotype analysis revealed that in sera primarily IgG was found, whereas in BAL both IgG and IgA were detectable. These antibodies were sufficient to prevent repeated gene transfer. The blocking humoral responses depended on class II presentation of viral proteins and activation of T-helper cells finally resulting in the formation of neutralizing antibodies.14 Twenty-eight days after infection with recombinant adenovirus via airways, BAL samples from mice displayed neutralizing activity. This neutralizing effect increased parallel to the rise in IgA antibodies, but was independent of the presence of IgG antibodies.20 The limited effect IgG from BAL has on inhibiting adenoviral infection as seen in our data was also demonstrated by Kaplan et al.7 In his study, BAL from adenovirus-infected rhesus monkeys retained neutralizing activity even after depletion of IgG antibodies. While IgG does not contribute to inhibition of adenoviral infection, its role in response to adenovirus remains unclear. IgG is mainly found in the lower respiratory tract with no increase after local nasal adenovirus application.21 Hypothesis was that IgG might be produced in regional lymph nodes and reach lung parenchyma by diffusion. Since IgG is known to be involved in opsonization to facilitate phagocytosis, it may be speculated that this process is not effective in eliminating adenovirus from mucosal surfaces. The effect of a single dose of adenovirus application on systemic antibody titer remains controversial. Crystal et al3 found no neutralizing antibodies in sera after recombinant adenovirus-mediated gene transfer to the human lung. The lack of systemic response may be contributed to by a low virus concentration and the fact that recombinant adenovirus does not replicate. Data about BAL findings were not reported. One year after a single application of adenovirus to the nose of rhesus monkeys, no antibody titer or neutralizing activity was detectable in nasal washes.7 In mice, serum IgA raised to a low level after a first administration of adenovirus vector to the nose.21 Antibody titers increased after repeated intranasal administration. Antibody titers found in this study were significantly higher than those described before, differing by one order of magnitude.12,22 The potential of antibodies in BAL or sera to neutralize adenovirus is subgroup-specific, although each subgroup demonstrates a low level of antigenic similarity. The major antigenic proteins are hexon, penton and fiber. The type-specific antigenic epitopes are located on fiber and hexons, neutralizing antibodies are directed mainly against hexons.23 Since epitopes on hexons cross-reacting between the serotypes are directed mainly towards the inside of the virion, infection with an adenovirus of one serotype does not produce relevant immunity against other types.24 Therefore, repeated application of adenovirus-mediated gene transfer using different serotypes seems feasible. However, in human studies previous infections with wild-type adenovirus may significantly reduce efficiency of gene transfer due to preformed IgA antibodies. In this regard previous adenovirus type 5 or 2 infections are of particular interest since most recombinant adenovirus vectors are derived from these types. Our results do not support the hypothesis that inhibitory properties of BAL are mediated just by immunoglobulins, such as IgA. Inhibitory effects may be caused by other components, such as proteases, proteins binding adenovirus unspecifically, or molecules competing with adenovirus for cellular binding sites. Possible candidates are fibronectin and vitronectin which are both present in Inhibition of adenovirus-mediated gene transfer by BAL A Bastian and B Bewig BAL and which inhibit uptake of adenovirus into cells.25,26 Like adenoviral penton base these molecules have RGD motifs involved in adenovirus internalization.27 Diseases associated with increased levels of inhibiting molecules in BAL (eg fibronectin in interstitial lung diseases) may therefore be less favorable for adenovirus-mediated gene therapy. In our study population, four of the patients whose BAL samples inhibited adenoviral infection most powerfully had sarcoidosis. However, the total number of patients was too small for this observation to be significant. Further studies will be necessary to detect components other than immunoglobulins in BAL which influence adenoviral infectivity. In this regard differences in diseases and in treatment, eg the application of steroids to alter the composition of the protein profile in BAL, may be important (eg effects on fibronectin or vitronectin). Although no in vitro scenario can simulate physiologic phenomena in vivo, we suggest pretesting BAL fluid from those patients who are to be treated for pulmonary diseases in gene therapy protocols using recombinant adenovirus. An assay for neutralizing activity may have significant implications for the treatment, since an inhibitory factor of 10 means that the dose necessary to be applied for equivalent effect needs to be increased 10-fold, possibly resulting in enhanced toxic side-effects. Materials and methods Subjects Twenty-six patients who underwent diagnostic bronchoscopy and bronchoalveolar lavage for sarcoidosis (9), tuberculosis (2), carcinoma of the lung (4), non-Hodgkin lymphoma (2), asthma (2), lymphocytic pneumonia (1), bronchitis (1) or exclusion of pulmonary disease (5) were randomly selected (Table 1). Bronchoalveolar lavage (BAL) Bronchoscopy was done using flexible fiberoptic bronchoscopes (Olympus, Hamburg, Germany). Patients received atropine 0.5 mg and hydrocodon 7.5 mg subcutaneously half an hour before examination. Up to 10 ml 1% oxybuprocainhydrochloride (Novesine; Wander, Nürnberg, Germany) or 2% lidocainehydrochloride (Xylocain; Astra, Hamburg, Germany) were used as local anesthesia. Supplemental oxygen was administered at a rate of 2 to 6 l/min. BAL was performed in the middle lobe bronchus or the bronchial segment affected by radiological changes. 200 ml of 0.9% sodium chloride solution was applied in 20 ml fractions, each fraction was aspirated by gentle suction and pooled in polypropylene vessels for further examination. Lavage fluid was filtered through sterile gauze and centrifuged at 500 g (10 min, 4°C). 50–100 ml of supernatant free of cells was applied to a centriprep 10 (Amicon, Beverly, MA, USA) and spun at 3000 g to a final volume of 500–700 l. After concentration, protein content was determined and adjusted with phosphate buffered saline (PBS) to a concentration of 1 g protein/l fluid. Quantification of adenoviral infectivity 293 Cells, a cell line derived from human embryonic nephroma, were seeded on 96-well culture dishes (Nunc, Wiesbaden, Germany) and grown to a density of 1 × 105 per well in 100 l of DMEM culture media (GIBCO BRL, Grand Island, NY, USA) containing 10% fetal calf serum (FCS), 100 U/l penicillin and 100 g/ml streptomycin. Cells were incubated with 20 l BAL (corresponding 20 g protein) or normal saline as control for 15 min. Recombinant adenovirus containing an empty expression cassette (AdCMV.Null) was prepared as previously described.28 10 l of AdCMV.Null in serum-free DMEM were applied to the 293 cells yielding a final concentration of 0.001 p.f.u. per cell. 1.5 h later complete medium, containing bovine serum at a concentration of 10% was added. Incubation was continued for additional 48 h. Media were aspirated and cells were allowed to dry in a laminar flow hood. Ice-cold methanol (100%) was added followed by an incubation at −20°C for 15 min. Methanol was aspirated and cells were washed twice with PBS containing 1% BSA (PBS/BSA). The following steps were performed at room temperature. A polyclonal rabbit antibody directed against adenoviral antigens was added to the cells (diluted 1:2000 in PBS/BSA, 50 l per well). After incubation for 1 h, cells were washed three times in PBS/BSA. A goat anti-rabbit antibody labeled with horseradish peroxidase was added (diluted 1:200 in PBS/BSA, 50 l per well) followed by another wash step (three times PBS). Infected cells were detected using the Pure Metal Enhanced Substrate Kit (Pierce, Rockford, IL, USA) which was applied according to the instructions of the manufacturer. After development of color, substrate was replaced by 200 l of PBS. All colored cell spots in each well were counted microscopically. Each experiment was done twice and as duplicate. Protein assay Protein concentrations were determined using the micro BCA protein assay reagent kit (Pierce) following the instructions of the manufacturer. The assay is based on the biuret reaction and adapted to microtiter plate usage. Briefly, 2 l of protein standards plus 148 l normal saline, blanks or samples were added to the plate followed by the application of 150 l of working solution (containing sodium carbonate, sodium bicarbonate and sodium tartate in NaOH, bicinchoninic acid in water and cupric sulfate and pentahydrate in water). After mixing and 120 min of incubation at 37°C absorbance was taken at 540 nm. Protein concentration was determined using a standard curve. Detection of IgA IgM and IgG anti-adenovirus antibodies Adenovirus stock solution was diluted 1:1000 in carbonate buffer (1.59 g Na2CO3 and 2.93 g NaHCO3 in 1 liter H2O, pH adjusted to 9.6). 75 l of this solution was applied to each well of a 96-well plate (Nunc, Wiesbaden, Germany) and incubated at 37°C in a humidified chamber overnight. Plates were washed three times with 0.05% Tween in PBS (PBS-Tween). 75 l of 2% BSA in PBS were added to each well and incubated at room temperature for 3 h. After three washings with PBS-Tween BAL samples containing 90 g of protein in 100 l NaCl were added. Incubation was performed for 4 h at 37°C. Washing was performed three times with PBS-Tween. 75 l of a 1:4000 dilution of goat anti-IgA antibody labeled with horseradish peroxidase (DAKO, Hamburg, Germany; No. PO216) were applied to the wells and incubated for 1 h at 37°C. 10 ml of substrate (0.1 m NaAc, 641 Inhibition of adenovirus-mediated gene transfer by BAL A Bastian and B Bewig 642 0.05 m NaH2PO4, adjusted to pH 4.2 with acetic acid, 0.5 mm 2,2′-azino-di-(3-ethylbenzthiazoline sulfonate) (ABTS), Boehringer, Mannheim, Germany) were mixed with 30 l of 3% H2O2. 100 l of this mixture were added to each well and absorbance was taken at 405 nm on an ELISA-plate reader. A standard curve was produced using dilutions of samples obtained from a patient with high concentrations of IgA. A 1:4000 dilution of goat anti-IgG antibody labeled with horseradish peroxidase (DAKO; No. PO214) and a 1:6000 dilution of goat anti-IgM antibody labeled with horseradish peroxidase (DAKO; No. PO215) was used, to determine IgG and IgM anti-adenovirus antibodies. Statistical analysis Factor of inhibition and content of immunoglobulin in BAL were analyzed to determine the correlation (r). Acknowledgements We would like to thank RG Crystal, Division of Pulmonary and Critical Care Medicine, CUMC, New York, for providing the adenoviral vector AdCMV.Null; E FalckPedersen, Department of Microbiology, CUMC, New York, for providing the rabbit anti-adenovirus antibody; and Sonja Rohweder for her excellent technical assistance. References 1 Ishibashi S et al. Hypercholesterolemia in low density lipoprotein receptor kockout mice and its reversal by adenovirusmediated gene delivery. J Clin Invest 1993; 92: 883–893. 2 Zabner J et al. Adenovirus-mediated gene transfer transiently corrects the chloride transport defect in nasal epithelial of patients with cystic fibrosis. Cell 1993; 75: 207–216. 3 Crystal RG et al. Administration of an adenovirus containing the human CFTR cDNA to the respiratory tract of individuals with cystic fibrosis. Nat Genet 1994; 8: 42–51. 4 Kay M. In vivo hepatic gene therapy: complete albeit transient correction of factor IX deficiency in hemophilia B dogs. Proc Natl Acad Sci USA 1994; 91: 2353–2357. 5 Yang Y et al. Cellular immunity to viral antigens limits E1deleted adenoviruses for gene therapy. Proc Natl Acad Sci USA 1994; 91: 4407–4411. 6 Yang Y, Xiang Z, Ertl HCJ, Wilson JM. Upregulation of class I major histocompatibility complex antigens by interferon ␥ is necessary for T cell-mediated elimination of recombinant adenovirus-infected hepatocytes in vivo. Proc Natl Acad Sci USA 1995; 92: 7257–7261. 7 Kaplan JM et al. Humoral and cellular immune responses of nonhuman primates to long-term repeated lung exposure to Ad2/CFTR-2. Gene Therapy 1996; 3: 117–127. 8 Yang Y et al. Inactivation of E2a in recombinant adenoviruses improves the prospect for gene therapy in cystic fibrosis. Nat Genet 1994; 7: 362–369. 9 Goldman MJ, Litzky LA, Engelhardt JF, Wilson JM. Transfer of the CFTR gene to the lung of nonhuman primates with E1deleted, E2a-defective recombinant adenoviruses: a preclinical toxicology study. Hum Gene Ther 1995; 6: 839–851. 10 Worgoll S, Wolff G, Falck-Pedersen E, Crystal RG. Innate immune mechanisms dominate elimination of adenoviral vectors following in vivo administration. Hum Gene Ther 1997; 8: 37–44. 11 Rosenfeld MA et al. In vivo transfer of the human cystic fibrosis transmembrane conductance regulator gene to the airway epithelium. Cell 1992; 68: 143–155. 12 Zabner J et al. Safety and efficiency of repetitive adenovirusmediated transfer of CFTR cDNA to airway epithelia of primates and cotton rats. Nat Genet 1994; 6: 75–83. 13 Kay MA et al. Long-term hepatic adenovirus-mediated gene expression in mice following CTLA41g administration. Nat Genet 1995; 11: 191–197. 14 Yang Y, Li Q, Ertl HCJ, Wilson JM. Cellular and humoral immune responses to viral antigens create barriers to lungdirected gene therapy with recombinant adenoviruses. J Virol 1995; 69: 2004–2015. 15 Kass-Eisler A et al. The impact of developmental stage, route of administration and the immune system on adenovirus-mediated gene transfer. Gene Therapy 1993; 1: 395–402. 16 Dai Y et al. Cellular and humoral immune response to adenoviral vectors containing factor IX gene: tolerization of factor IX and vector antigens allows for long-term expression. Proc Natl Acad Sci USA 1995; 92: 1401–1405. 17 Zsengeller ZK et al. Persistence of replication-deficient adenovirus-mediated gene transfer in lungs of immune-deficient (nu/nu) mice. Hum Gene Ther 1995; 6: 457–467. 18 Ginsberg HS et al. Role of early region E (E3) in pathogenesis of adenovirus disease. Proc Natl Acad Sci USA 1989; 86: 3823–3827. 19 Vilquin JT et al. FK506 immunosuppression to control the immune reactions triggered by first-generation adenovirusmediated gene transfer. Hum Gene Ther 1995; 6: 1391–1401. 20 Yang Y, Trinchieri G, Wilson JM. Recombinant IL-12 prevents formation of blocking IgA antibodies to recombinant adenovirus and allows repeated gene therapy to mouse lung. Nature Med 1995; 1: 890–893. 21 Van Ginkel FW et al. Intratracheal gene delivery with adenoviral vector induces elevated systemic IgG and mucosal IgA antibodies to adenovirus and -galactosidase. Hum Gene Ther 1995; 6: 895–903. 22 Brody SL et al. Acute responses of non-human primates to airway delivery of an adenovirus vector containing the human cystic fibrosis transmembrane conductance regulator cDNA. Hum Gene Ther 1994; 5: 821–836. 23 Valentine RC, Pereira HG. Antigens and structure of the adenovirus. J Mol Biol 1965; 13: 13–20. 24 Mastrangeli A et al. ‘Sero-switch’ adenovirus-mediated in vivo gene transfer: circumvention of anti-adenovirus humoral immune defenses against repeat adenovirus vector administration by changing adenovirus serotype. Hum Gene Ther 1996; 7: 79–87. 25 Rennard SI, Crystal RG. Fibronectin in human bronchopulmonary lavage fluid. J Clin Invest 1981; 69: 113–122. 26 Teschler H et al. Elevated levels of bronchoalveolar lavage vitronectin in hypersensitivity pneumonitis. Am Rev Respir Dis 1993; 147: 332–337. 27 Wickham TJ, Mathias P, Cheresh DA, Nemerow GR. Integrins ␣v3 and ␣v5 promote adenovirus internalization but not virus attachment. Cell 1993; 73: 309–319. 28 Hersh J, Crystal RG, Bewig B. Modulation of gene expression after replication-deficient, recombinant adenovirus-mediated gene transfer by the product of a second adenovirus vector. Gene Therapy 1995; 2: 124–131.