Towards the In Vivo Identificaton of Leukemogenic Fusion Proteins.

Currently there are no methods available to identifiy leukemogenic fusion proteins in vivo. All available methods, like Southern blotting, PCR, FISH or Western blotting, require the destruction of the cells that are assayed. A method for the in vivo detection of leukemogenic fusion proteins would be highly desirable because it would open up new approaches to study leukemia and might lead to novel treatment strategies.

We have developed a strategy for the in vivo detection of the BCR/ABL fusion protein. BCR/ABL is found in virtually all cases chronic myeloid leukemia (CML) and a large proportion of acute lymphoblastic leukemia (ALL). Animal model have shown that the BCR/ABL fusion protein is required for the induction and maintenance of leukemia. The fact that BCR/ABL fusion protein is crucial for the development of leukemia makes this fusion protein an attractive target for therapy development. Our BCR/ABL detection strategy is based on protein-protein interactions and a proof of principle for the strategy was implemented in the yeast system. Two detection proteins are expressed in the cells: 1) protein A, a Gal4-DNA binding domain/BCR interacting protein fusion protein and 2) protein B, a Gal4-activation domain/ABL interacting protein fusion protein. Only when BCR/ABL is present in the cell, do protein A, protein B, and BCR/ABL form a trimeric complex which activates the transcription of reporter genes under the control of Gal4-upstream activating sequence (UAS).

Yeast cells (strain CG1945) transformed with a protein A expressing plasmid (pGBT9-BCR-interactor), a protein B expressing plasmid (pGAD424-ABL1-interactor), and a BCR/ABL expressing plasmid (pES1-BCR/ABL) showed expression of the reporter genes HIS3 and LACZ. The expression of the HIS3 reporter gene was assayed by growth of the yeast cells on medium lacking histidine. The expression of the LACZ gene was verified by a beta-galactosidase filter assay.

Yeast cells that were transformed with the pES1 plasmid without the BCR-ABL coding region did not show activation of the reporter genes. Several other negative controls were also negative. Thus the method was able to clearly distinguish between BCR/ABL expressing cells and cells did not express BCR/ABL.

We are presently adapting this system for use in mammalian cells. The flexibility of our strategy allows us to freely choose the reporter or effector genes. Therapeutically more useful effector genes are suicide genes, which encode pro-drug converting enzymes (e.g. HSV thymidine kinase), or markers that can easily be assayed (e.g. green fluorescent protein).