Conventional protein-protein interaction (PPI) mapping methods, such as affinity purification-mass spectrometry (AP-MS) and yeast-2-hybrid (Y2H) systems, often fail to capture low-abundance, transiently expressed, or membrane-bound proteins. This limitation arises from their inability to replicate the native physiological conditions of a living cell1,2,3. To overcome these challenges, PL has emerged as a powerful technique for high-resolution proteome mapping at a sub-organelle level in vivo. PL leverages the fusion of a promiscuous enzyme with bait protein, enabling the enzyme to catalyze the formation of short-lived reactive molecules1. These reactive molecules, such as biotin, covalently label proteins within a few nanometers of the fusion protein4. Subsequently, biotinylated proteins are isolated and analyzed by mass spectrometry to facilitate large-scale protein identification5,6.
Over the past years, various promiscuous enzymes have been developed to advance PL techniques. Two widely used enzymes are Escherichia coli-derived R118G mutant biotin ligase (BioID) and pea-derived ascorbate peroxidase enzyme 2 (APEX2). BioID converts biotin and ATP into biotinyl-5'-adenylate (bioAMP)1. This reactive molecule covalently binds to lysine residues of nearby proteins within ± 10 nm radius. A key advantage of BioID-based PL is the use of biotin. It is a naturally occurring and non-toxic biomolecule in living organisms, which allows for safe in vivo labeling6. However, BioID has limitations, including slow labeling kinetics (18-24 h) and the requirement for high temperature (37 °C) for optimal catalytic activity4. These traits can hinder its application in studying dynamic biological processes. Meanwhile, APEX2 catalyzes the formation of biotin-phenoxyl radicals in the presence of biotin-phenol and hydrogen peroxide (H2O2)7. These radicals mainly react with the side chain of electron-rich amino acids like tyrosine and can also bind cysteine, histidine, and tryptophan8. The rapid labeling capability of the peroxidase enzyme (< 1 min) is suitable for capturing dynamic and transient PPIs9. Moreover, it has a wider detection range (up to 20 nm) than biotin ligase. However, APEX2-based PL has its drawbacks, including the requirement for a toxic oxidizing agent, H2O2, and the low permeability of biotin-phenol, which limit the applicability in vivo10.
In 2018, Branon et al. developed a new biotin ligase called TurboID, a 35 kD mutant version of BirA (the biotin ligase found in E. coli)4. This enzyme incorporates a mutation at the R118 position (R118S) together with 15 other mutations relative to BirA. It exhibits two-fold higher catalytic activity than BioID. TurboID-induced biotinylation in 10 min produces proteomic data with a size and specificity comparable to 18 h of labeling by BioID. Furthermore, it demonstrates excellent labeling activity at 30 °C4. Therefore, TurboID is more suitable for application in organisms like Drosophila or Caenorhabditis elegans, which are commonly reared at 25 °C and 20 °C, respectively.
In this study, we aim to utilize TurboID-based PL for interactome study in the Drosophila ovary. The protocols described here are employed to map the interactome of Zuc in germline cells. Zuc is an endonuclease localized on the outer mitochondrial membrane (OMM), which mediates the biogenesis of piRNAs, small non-coding RNAs critical for the maintenance of genome integrity11,12,13. By comparing biotinylated proteins identified from Zuc-TurboID analysis with those from two additional controls, NES-TurboID and Tom20-TurboID, we defined distinct Zuc-interacting candidates.