In vivo dose threshold effect of adenovirus-mediated factor VIII gene therapy in hemophiliac mice.

While much is known about adenovirus biology from its development as a therapeutic gene delivery vehicle, an important question remains regarding the appropriate in vivo vector dose. We describe here an in vivo dose threshold effect with an adenoviral vector expressing human Factor VIII (FVIII) in hemophiliac mice. Upon administration of vector doses between 6 x 10(10) and 2 x 10(10) vector particles per mouse, FVIII was expressed linearly, whereas a dose of 1 x 10(10) vector particles per mouse did not result in detectable levels of FVIII activity. In contrast, in vitro transduction studies demonstrated linear transgene expression over 2 to 3 log units. To further define this dose threshold effect, a vector-mixing study was performed. Mice were injected with a total vector dose of 6 x 10(10) particles containing admixtures of FVIII vector plus a control vector lacking a transgene (null vector). With the admixture, FVIII activity was detected in mice that received 1 3 1010 particles of the FVIII vector, indicating that maintenance of the total viral input at 6 x 10(10) particles per mouse circumvented the threshold dose effect. This threshold dose effect could not be attributed to dose-dependent differences in liver toxicity nor to dose-dependent induction of cellular and humoral immune responses. Southern blot analysis of livers revealed that mice receiving the vector admixture contained FVIII DNA, accounting for the observed FVIII expression, whereas mice receiving 1 x 10(10) particles of FVIII vector had barely detectable FVIII DNA. These results suggest that the threshold effect is an in vivo phenomenon that will have important implications in defining the therapeutic window of adenoviral vectors for clinical applications.