Developing New Enzymatic Scaffolds for Proximity Labeling

Unbiased, proteomic approaches for defining the molecular components of a spatially defined region of interest often rely on long extensive purification schemes to isolate the region of interest. Proximity labeling based proteomic approaches have revolutionized this workflow for molecular discovery because they circumvent the need for classical forms of biochemical fractionation, allowing access to many previously unpurifiable subcellular structures. Most iterations of proximity labeling rely on the expression of an enzyme in the region of interest to generate a reactive intermediate that tags neighboring proteins with a biotin handle, for subsequent streptavidin enrichment and identification by proteomics. Although applied widely and successfully in vitro and in cell culture systems, proximity labeling suffers from several limitations when applied in vivo. The major limitation when applied in vivo arises from additional sources of background in tissue, particularly from endogenous biotinylated proteins that similarly compete for the same binding site during streptavidin enrichment. This limitation is particularly debilitating when the region of interest is a small subpopulation of cells within the tissue, such as a small subpopulation of neurons within the brain. This challenge could, however, be overcome by using non-biotin-streptavidin based approaches for chemical enrichment of labeled proteins. As a step towards expanding the in vivo applications of proximity labeling, this proposal will identify or engineer new enzyme scaffolds, capable of generating reactive intermediates like those generated by the proximity labeling enzyme TurboID, that have been functionalized for chemical enrichment instead of biotin-streptavidin enrichment.