Conjugation of an antibody Fv fragment to a virus coat protein: cell-specific targeting of recombinant polyoma-virus-like particles.

The development of cell-type-specific delivery systems is highly desirable for gene-therapeutic applications. Current virus-based vector systems show broad cell specificity, which results in the need to restrict the natural tropism of these viral systems. Here we demonstrate that tumour-cell-specific virus-like particles can be functionally assembled in vitro from recombinant viral coat protein expressed in Escherichia coli. The insertion of a negatively charged peptide in the HI loop of polyoma VP1 interferes with the binding of VP1 to the natural recognition site on mammalian cells and also serves as an adapter for the coupling of antibody fragments that contain complementary charged fusion peptides. A recombinant antibody fragment of the tumour-specific anti-(Lewis Y) antibody B3 could be coupled to the mutant VP1 by engineered polyionic peptides and an additional disulphide bond. With this system an entirely recombinant cell-specific delivery system assembled in vitro could be generated that transfers genes preferentially to cells presenting the tumour-specific antigen on the cell surface.

Coupling of antibodies via protein Z on modified polyoma virus-like particles.

Therapeutic application of virus-based delivery systems often implies a change of the tropism of these vectors. This can be achieved by insertion of polypeptides (e.g., antibody fragments) in viral coat proteins. Such fusion proteins have only been used in viral vectors so far and, as part of a virus, they have not been available for a detailed biophysical characterization. We analyzed a fusion protein called VP1-Z, which is based on the polyoma virus coat protein VP1 and protein Z. Protein Z is an engineered antibody-binding domain derived from protein A from Staphylococcus aureus. The fusion VP1-Z was constructed by insertion of protein Z in the HI-loop of VP1. As wild-type VP1, VP1-Z formed pentameric capsomers and assembled to VLPs in vitro. The stability of these particles was very similar compared to that of VLPs of wild-type VP1. Protein Z was fully structured in the fusion protein and was still capable of binding antibodies on the surface of VLPs of VP1-Z. Using this fusion protein, we could change the tropism of polyoma VLPs toward cells presenting on their surface the antigen of the coupled antibody.

Changing the surface of a virus shell fusion of an enzyme to polyoma VP1.

Recent developments on virus-like particles have demonstrated their potential in transfecting eucaryotic cells. In the case of particles based on the major coat protein VP1 of polyoma virus, transfection occurs via binding of VP1 to sialic acids. Since sialic acid is present on almost every eucaryotic cell line, this results in an unspecific cell targeting. Generation of a cell-type specificity of this system would imply the presentation of a new function on the surface of VP1. To analyze whether a new functional protein can be placed on VP1, we inserted dihydrofolate reductase from Escherichia coli as a model protein. The effect of such an insertion on both VP1 and the inserted protein was investigated, respectively. The function of VP1, like the formation of pentameric capsomers and its ability to assemble into capsids, was not influenced by the insertion. The inserted dihydrofolate reductase showed major changes when compared to the wild-type form. The thermal stability of the enzyme was dramatically reduced in the fusion protein; nevertheless, the dihydrofolate reductase proved to be a fully active enzyme with only slightly increased K(M) values for its substrates. This model system provides the basis for further modifications of the VP1 protein to achieve an altered surface of VP1 with new properties.