Preferential MyoD homodimer formation demonstrated by a general method of dominant negative mutation employing fusion with a lysosomal protease

We report on a general strategy for engineering dominant negative mutations that, in principle, requires neither extensive structural or functional knowledge of the targeted protein. The approach consists of fusing the lysosomal protease cathepsin B (CB) to a subunit of a multimeric protein. The CB fusion polypeptide can proteolytically digest the multimer and/or detour the multimer from its usual subcellular destination to the lysosome. We first demonstrate the general validity of the approach with CB fusion to E. coli lacZ, encoding tetrameric beta-galactosidase. Cotransfection of NIH 3T3 cells with a vector expressing a CB-lacZ fusion inhibits the beta- galactosidase activity produced by transfection of lacZ alone. We infer that the dominant negative inhibition results from both direct proteolysis of the beta-galactosidase tetramer by the fusion subunit and detour of the tetramer to the lysosome. In a specific application of this strategy, we have fused CB to the dimeric bHLH skeletal muscle transcription factor MyoD. The CB-MyoD fusion protein localizes to the cytoplasm, presumably the lysosome, demonstrating the dominance of lysosomal localization to nuclear localization. The CB-MyoD fusion appears to divert homodimerizing native MyoD from its usual nuclear destination, consequently inhibiting MyoD-mediated transactivation and in vitro differentiation of C2C12 myoblasts. Surprisingly, the CB-MyoD fusion fails to interact with the bHLH heterodimerization partners, E12 and E47, suggesting preferential MyoD homodimer formation, at least in the prenuclear cellular compartments.