The hepatotropic properties of human adenoviruses have been used to develop vectors for in vivo liver-directed gene therapy. Current limitations for this vector system are the associated hepatitis that develops as a result of antigen-specific cellular immune responses and the difficulty in accomplishing repeated gene transfer. This study uses mouse models to define immune responses of the recipient animal that have previously been shown to prevent successful re-administration of virus and suggests approaches for preventing the development of these blocking immune responses. Our studies are most consistent with class II MHC-dependent activation ot T helper cells and B cells to capsid proteins of the input virus leading to the production of antiviral neutralizing antibody following a primary exposure to virus; this capsid-specific antibody appears to bind to virus and prevents entry in the context of a second administration of virus. Transient ablation of CD4 function at the time of virus administration prevents the formation of neutralizing antibody thereby allowing efficient gene transfer after at least two subsequent administrations of virus. Experiments in beta(2)m(-) mice and C57BL/6 mice treated with IL-12 suggested a more selective ablation of immune function based on inhibiting the activation of the T(H2) subset of T helper cells. From these studies on immune mechanisms it is hoped that viable strategies can be developed to overcome the problem of humoral immunity that occurs after the initial genetic therapy.