Engineering viral vectors to subvert the airway defense response.

The potential treatment of lung disease by delivery of exogenous DNA to cells for expression of therapeutic proteins is an exciting benefit of the recent revolution in molecular engineering (1–3). Adenovirus-based vectors have received considerable attention for DNA delivery to lung cells for several reasons: adenoviruses can infect nondividing cells in the airway epithelium; they are stable; they can be produced in high titer; and they are easily modified to be replication-deficient and express exogenous genes. Adenoviral vectors have already demonstrated the capacity for gene transfer in cell culture systems (4, 5), animal models (4, 6, 7), and human trials (8–10). However, as development of this therapeutic technology progressed to in vivo systems, biologic obstacles to gene therapy have emerged that limit high-level and prolonged transgene expression (1–3). The antiviral defense response of the host is one obstacle that has particularly tempered enthusiasm for earlygeneration adenoviral vectors, although recent modifications of vectors or host immunity have had some success at inhibiting this response. Accordingly, the purpose of this perspective is to review briefly the host response to adenoviruses and to discuss potential strategies to overcome this obstacle to adenovirus-based gene therapy to the lung. Administration of adenoviral vectors into the lung generates a complex, multicomponent host response that involves innate and adaptive immunity, resulting in specific inhibitory effects on gene therapy efficacy (2). Components of the pulmonary response to adenovirus include: ( 1 ) an antigen-nonspecific, cytokine-dependent response resulting in acute inflammation, which may have a neurogenic component and can cause systemic toxicity (7, 9–12); ( 2 ) a major histocompatibility complex (MHC) class I– restricted, cytotoxic (CD8 1 ) T lymphocyte–dependent response directed at cells expressing viral or transgene proteins, resulting in chronic inflammation and a lack of persistent transgene expression (7, 11, 13); and ( 3 ) a helper (CD4 1 ) T lymphocyte–dependent response directed at adenoviral capsid proteins present during vector delivery, resulting in the production of neutralizing antibodies that limit repeated vector administration (if required to sustain a therapeutic effect) (10, 13, 14). It should not be surprising that the response to viral pathogens occurs through multiple, interrelated mechanisms because a multicomponent system can probably adapt for host-defense against a variety of pathogens. However, it is apparent that all components of the antiviral response in the lung will need to be identified and controlled in order to improve gene therapy efficacy with adenoviral vectors. In this issue, Thorne and colleagues (15) present a murine model in which high-dose exposure to inactivated adenoviral particles elicits an acute airway inflammatory response marked by interleukin (IL)-6 and tumor necrosis factora release and consequent neutrophil recruitment into the lung. This model appears relevant to gene therapy in humans because IL-6 release was observed after adenoviral vector administration to the lungs of patients with cystic fibrosis (CF) (9). The early pulmonary response to the adenovirus studied by Thorne and colleagues does not appear to be additive with the response to endotoxin, suggesting that both stimuli may activate inflammation by overlapping pathways. This finding has implications for adenovirus-based gene therapy in patients with preexisting inflammation from endotoxin, as this is part of the airway milieu in patients with lung diseases such as CF, characterized by chronic infection with gram-negative bacilli. This study also confirms that a portion of the host response to adenovirus is independent of vector gene expression, but may be modulated pharmacologically (16). Because this acute response appears to be independent of viral gene expression, future work will need to be directed toward identification of the initiating elements in adenoviral particles with the hope that these elements might be altered to limit the host response. As host responses to adenoviral vectors that limit the utility of adenovirus-based gene transfer have been identified, two major strategies to blunt these responses have been tested. The first strategy is formed on the basis of identifying specific viral molecules that initiate or amplify the pulmonary defense response with subsequent vector reengineering to delete their expression. Work in this direction spans from removal of specific viral gene products (e.g., early region 2A) to total removal of all genes in the vectors that express viral proteins (17, 18). A second strategy is to pharmacologically manipulate the host response using drugs (e.g., corticosteroids, cyclophosphamide), mediators (e.g., IL-12), or other blocking strategies (e.g., anticell receptor proteins, oral tolerization) (19–22). Both ( Received in original form April 2, 1999 )