(626a) Complementation Assay for Detecting Homodimeric Association On the Yeast Surface

Proteins in nature frequently form oligomers and therefore modulating protein-protein association can help understand their in vivo functions. In this regard, there are various assays for studying and optimizing protein association, including yeast two hybrid and several protein complementation assays. However, these assays are mostly used to study heterodimeric interactions, in which the interaction between two non-identical subunits is measured. In contrast, there are few convenient methods for detecting homodimerization in vivo. Since homodimers are the most common form of protein oligomers, detecting homodimerization should be a useful tool for studying protein function. A high throughput homodimerization assay would also allow new homo- and heterodimers to be engineered by directed evolution in ways that are difficult to implement using any other existing technique. To this end, we developed an assay to determine protein homodimerization on the yeast surface. The assay uses an engineered streptavidin monomer that selectively forms a domain-swapped dimer with high biotin affinity and is based on a recent discovery that streptavidin, which has always been considered an obligate tetramer, can exchange a biotin binding loop to form a functional homodimer. When this monomeric streptavidin is fused to a heterologous protein that self-associates, the homodimerization of the fusion protein results in reconstitution of a functional, domain-swapped streptavidin dimer, which can be detected based on biotin binding. Because domain swapping is inefficient, fusing a protein domain that self-associates promotes the creation of a functional streptavidin complex. We demonstrate this strategy using two leucine zipper proteins, one of which (Jun) homodimerizes with high affinity while the other (Fos) does not homodimerize.