Nouvelle approche pour modifier le tropisme des vecteurs adenoviraux a l'aide de ligands bispecifiques

Adenoviruses have been studied as a way to develop new treatments for different diseases. Adenoviral vectors (AdV) are considered interesting tools for this propose, because they can be produced at high titers (1X10 12 particles per millilitre) in laboratory and they have the capacity to infect non-dividing and dividing cells. AdV have been often modified in order to obtain the ability to kill tumour cells or to deliver exogenous genetic sequences essential to treat monogenic disease. However, weak expression of the primary adenovirus receptor, the CAR (Coxsackie and adenovirus receptor) reduces greatly the transduction efficiency of AdV for the tumour cells. Moreover, some normal tissues express low amount of CAR, like the skeletal muscle, reducing the appeal of using AdV as a gene delivery vehicle for this tissue. To address this problematic, many modifications were done on the adenoviral capsid. The goal of these modifications were to generate an AdV able to target specific cellular receptors that were expressed in tumour cells but not in normal cells. Several approaches were done to modify the tropism of AdV, such as incubation with a bispecific ligands, incorporation of peptides within the adenoviral fiber structure or substitution of the viral fiber with a different serotype fiber.;The hypothesis of my project was to determine if an interaction domain fused within a ligand could bind the complementary domain incorporated on a virus and change the tropism of the AdV. The first step was to include a synthetic interaction domain, the K-Coil, within specific region of the adenoviral fiber, as well as inserting two point mutations to abolish the natural tropism. To target the EGF-R, IGF-IR and the CEA6, we fused the complementary interaction domain, the E-Coil, to the respective ligand known as the EGF and the IGF-I or to a single domain antibody (known as AFAI) that bind specifically to CEA6. The specific interaction between the E-Coil and K-Coil was used to associate the ligand with the fiber in order to retarget the AdV toward the selected receptor.;We showed that the different ligands as well as the modified fibers could be produced and that both E-Coil and K-Coil expressing partners could interact together. We optimized the viral production by using an iodixanol purification protocol. More importantly, we clearly demonstrated that the ligand association with the fiber could increase the transduction efficiency between 2 to 21 fold against various tumour cells. The difficulty of adenovirus to infect muscle cells because of the lack of CAR expression brought us to evaluate the potential of our retargeted AdV to increase the transduction for the tissue. We showed that the use of IGF-E5 could increase the transduction efficiency in myoblasts as wells as in myotubes. We finally demonstrated that our retargeting system could increase the transduction efficiency for skeletal muscle by 1,6 fold in new born MDX mice. In conclusion, our results show that the retargeting system is indeed functional. This system could be assessed using vectors that express therapeutic genes.;Key words: Adenovirus, Retargeting, IGF-IR, Tumours, Muscles.