45. Adenovirus Variant With a Defined Set of Genetic Mutations in the Penton and Hexon Escapes Liver Sequestration After Intravascular Administration

Abstract

Adenoviruses (Ad) are promising vectors for therapeutic interventions in humans. Despite significant knowledge regarding the biology of Ad interactions with cells in vitro, the molecular mechanisms governing in vivo Ad infectivity and bio-distribution remain poorly understood. Pharmacokinetic studies of Ad vectors after intravascular delivery demonstrate that the majority of an administered virus dose is rapidly sequestered from the circulation by the liver. Earlier we showed that Ad particles are distributed in liver tissue amongst three distinct cellular compartments, namely i) parenchymal liver cells – hepatocytes, ii) hepatic residential macrophages, Kupffer cells, and iii) hepatic sinusoid endothelial cells. Importantly, we found that the ablation of Ad interaction with only one of these cellular compartments cannot prevent virus trapping the liver shortly after intravascular virus injection and, likely, results in compensatory re-distribution of the virus among two remaining cellular compartments, thus functionally ensuring the quantitative removal of the virus from the blood. We speculated that to prevent Ad sequestration by the liver after intravascular administration, simultaneous ablation of virus interaction with all three cellular compartments of the liver is ultimately required. To evaluate this assumption experimentally, we introduced in a single human Ad5-based vector specific mutations that abrogate virus interactions with one, two, or all three hepatocellular compartments. To ablate Ad interaction with hepatocytes in vivo, we introduced T425A mutation in hexon hyper-variable loop HVR7 that completely abrogates virus interaction with blood coagulation FX. To prevent Ad interaction with liver sinusoid endothelial cells, we introduced a substitution of the RGD-motif-containing penton loop for an iso-functional integrin-interacting non-RGD-containing peptide. We also serendipitously found that the deletion of hexon HVR1 region, in addition to penton RGD-loop substitution and T425A HVR7 mutation, results in the reduction of virus interaction with Kupffer cells. Intravenous injection into mice of Ad variants containing individual mutations either in the penton or hexon or a vector with only two mutations in a single vector does not prevent sequestration of resultant Ad in the liver. However, Ad variant, containing all three mutations simultaneously, completely escaped being sequestered in the liver after intravascular injection and, instead, 20-fold higher amounts of the virus were recovered from the spleen, indicating the major change in virus bio-distribution, compared to unmodified vectors with a wild-type capsid, or vectors with individual mutations in either penton or hexon. To our knowledge this is the first demonstration of a feasibility of constructing Ad variants that would escape liver sequestration after intravascular administration. These variants may prove to be a useful platform for gene delivery to extra-hepatic cells and/or targeting disseminated metastatic cancers.