15. Genetically Targeted Adenovirus Vector Carrying Reovirus σ1-Based Attachment Protein for Improved Systemic Gene Delivery

Abstract

In vivo applicability of adenoviral vectors (AdV) for cancer gene therapy is limited by sequestration of AdV by the liver and lack of adenovirus receptor CAR on many primary tumors. To circumvent these limitations we developed prototype genetically targeted AdV that lack all known adenovirus interaction sites with cellular receptors (CAR, HSG and integrins) and carry a unique target-binding moiety. The adenovirus fiber protein was replaced with a chimeric molecule comprising the capsid-anchoring tail domain of fiber, the oligomerization T(ii) domain of reovirus attachment protein |[sigma]|1 and a His-tag as model targeting moiety. Furthermore, the RGD motif in the penton base protein was mutated. The chimeric |[sigma]|1-based molecule was incorporated into AdV capsids and allowed efficient propagation of AdV without requirement for complementing fiber. Importantly, the His-tag conferred targeted tropism to the AdV, resulting in up to 50-fold greater transduction efficiency on cells expressing a His-tag binding receptor than on parental cells lacking this receptor. Targeting was mediated through binding of the chimeric molecule, as transduction was strongly inhibited by a His-tag specific antibody. In focus of the important issue of liver de-targeting, we studied transduction of HepG2 cells and mouse liver cells in explanted liver slices with intact tissue architecture. In comparison to control AdV with native tropism, genetically targeted AdV exhibited 3-log reduction in HepG2 transduction and 3-4-log reduction in the stringent liver slice model. We also studied in vivo persistence and biodistribution of genetically targeted AdV after systemic delivery by tail vein injection into C57bl/6 mice. Blood samples collected during the first 2 hours after injection revealed rapid clearance of the control vector with native tropism, leaving less than 1% of the administered dose after 10 minutes, which declined further to less than 0.1% after 30 minutes. In contrast, the targeted AdV declined to 1% of the administered dose only after 1 hour, and this level remained stable for at least the next hour. Analysis of all tested tissues, i.e., liver, heart, spleen, lungs and kidney revealed a significant reduction of transduction by the targeted AdV in comparison to the control vector. In fact, transduction of lung and kidney by the targeted AdV was undetectable. Finally, in contrast to AdV with native tropism, the targeted AdV did not agglutinate human red blood cells. This suggests that the targeted AdV would be less susceptible to sequestration by red blood cells in the human circulation. Together, our findings suggest that the reovirus |[sigma]|1-based chimeric attachment molecule provides a promising platform for generation of targeted AdV with improved characteristics for systemic in vivo gene delivery.