Directed evolution of a probe ligase with activity in the secretory pathway and application to imaging intercellular protein-protein interactions

Katharine A White,Phillip M Zegelbone
DOI: https://doi.org/10.1021/bi400268m
IF: 3.321
2013-05-28
Biochemistry
Abstract:Previously, we reported a new method for intracellular protein labeling in living cells called PRIME (probe incorporation mediated by enzymes). PRIME uses a mutant of Escherichia coli lipoic acid ligase (LplA) to catalyze covalent probe ligation onto a 13-amino acid peptide recognition sequence. While our first demonstration labeled proteins with a coumarin fluorophore, subsequent engineering produced alkyl azide and trans-cyclooctene ligases as well as an interaction-dependent form of the coumarin PRIME method (ID-PRIME). One major limitation of the PRIME methodologies is that LplA mutants have very low activity in the secretory pathway. Here, we extend PRIME labeling to oxidizing compartments such as the endoplasmic reticulum and the cell surface. We used yeast-display evolution and four rounds of selection to isolate LplA mutants with improved picolyl azide ligation activity. Then we compared the ligation activities of the evolved mutants both in vitro and on the mammalian cell surface. We characterized the picolyl azide ligation activity of the most active LplA variant in vitro, in the endoplasmic reticulum, and at the mammalian cell surface. Finally, we used the optimized LplA variant to label neurexin and neuroligin interactions at the mammalian cell surface in just 5 min. Compared to another method for imaging these protein-protein interactions (GFP recomplementation across synapses), our optimized ID-PRIME ligase is faster, more sensitive, and does not trap interacting proteins in a complex (nontrapping).
What problem does this paper attempt to address?