Current vectors that are being engineered include adenovirus, adeno-associated virus, herpes virus, and non-viral vectors. The goal of our efforts is to develop generic technologies that can target any cell type using one or more vector platforms. One example is a technology we have developed called metabolic biotinylation. When applied for vector targeting, viral proteins are genetically tagged with biotin acceptor peptides (BAPs). Therefore, any biotinylated ligand can in theory be used to target vectors. This also means that the vector needs to be genetically engineered only once, avoiding the necessity to engineer a new vector for each new ligand. Finally, structural compatibility problems are avoided between ligands and vector proteins such that large ligands (e.g. antibodies) and non-protein ligands (e.g. carbohydrate ligands, lectins, DNA, etc.) can now be combined with a vector by one common conjugation technology
Metabolic Biotinylation Application: Screening Viral Capsomers as Scaffolds for Vector Targeting and Purification. The direct genetic modification of adenoviral capsid proteins with new ligands is an attractive means to confer targeted tropism to adenoviral vectors. Although several capsid proteins have been reported to tolerate the genetic fusion of foreign peptides and proteins, direct comparison of cell targeting efficiencies through the different capsomeres has been lacking. Likewise, direct comparison of with one or multiple ligands has not been performed due to a lack of capsid-compatible ligands available for retargeting. Given metabolic biotinylation allows one to tag a number of capsomer proteins with BAPs and then combine these with any biotinylated ligand, this provides a unique platform to compare the functions of a given viruses proteins for targeting and purification.
As a first example of this approach, we have utilized metabolic biotinylation to directly compare targeted transduction of adenovirus as mediated by its fiber, protein IX, and hexon capsomeres using a variety of biotinylated ligands (Virology 349(2):***********). These results clearly demonstrate that cell targeting with a variety of high affinity receptor-binding ligands is only effective when transduction is redirected through the fiber protein. In contrast, protein IX and hexon-mediated targeting by the same set of ligands failed to mediate robust vector targeting, perhaps due to aberrant trafficking at the cell surface or inside targeted cells. These data suggest that vector targeting by genetic incorporation of high affinity ligands will likely be most efficient through modification of the adenovirus fiber rather than the protein IX and hexon capsomeres. In contrast, single-step monomeric avidin affinity purification of Ad vectors using the metabolic biotinylation system is most effective through capsomeres like protein IX and hexon. We predict the same rules may apply when targeting ligands are applied on other virus proteins that are not evolved to release from the virion during cell entry. Likewise, we predict chemical conjugation approaches that cross-link ligands to non-shedded proteins may also confront these problems. Work is underway to determine exactly how these modified vectors fail, how the affinity of the ligands affects this process, and how one might use or avoid this biology to maximize vector specificity.
Metabolic Biotinylation Application: Biotinylated Vectors as Ligand Screening Platforms for Ligands. Metabolic biotinylation allows one genetically engineered vector to be used with any biotinylated ligand. As such, biotinylated vectors can be used as a platform for combinatorial screening of both cell-targeting ligands and for receptors that are compatible with cell targeting. As proof of principle for this application, we have screened for useful ligands to increase transduction of primary dendritic cells (Molecular Therapy 8(4): 689-702 2003). The following ligands were tested using the biotinylated fiber vector: 1) biotinylated antibodies against CD40, CD86, CD11b, CD11c, MHC II, 2) biotinylated protein ligands targeting TLR receptors and scavenger receptors (HSC70, aggregated albumin), biotinylated mannosylated BSA targeting the mannose receptor, biotinylated asialofetuin targeting the dendritic asialofetuin receptor, and biotinylated CpG oligonucleotides to target TLR9 (Fig. 3). These diverse and structurally incompatible ligands were easily screened when complexed to biotinylated adenovirus using avidin as a bridge. Work is ongoing to use this technology to prescreen ligands on our varied targeting models.
Metabolic Biotinylation Application: Biotin Tags as Cryo-EM Tags. Metabolic biotinylation allows one genetically tag any capsomer of a virus with a small, but detectable electron density that can be observed by imaging. As proof of principle for this approach, we have tagged the IX protein of adenovirus with a 70 amino acid BAP and performed cryo-EM of this virus. This work in collaboration with Mike Marsh and Wah Chiu at Baylor College of Medicine led to the reassignment of the location of the IX protein (J. Virol. 80: 11881-11886 2006). Rather than being positioned in the 79 Å valleys between hexon trimers on the icosahedron, IX appears to be located on the surface of the virion between the H2 hexon and the H4 hexon, positioned between adjacent facets, directly above the density previously assigned as protein IIIa. The original assignment of IIIa was based largely on indirect evidence and the data generated by BAP and other molecular tagging of capsomers support the reassignment of the IIIa density as protein IX. Work is underway to apply this approach for other capsomers and to apply cryo-EM as an additional method to better understand the biology of native and engineerd cell-targeting viruses.