Comparative Application of BioID and TurboID for Protein-Proximity Biotinylation

Abstract

BioID is a well-established method for identifying protein-protein interactions and has been utilized within live cells and several animal models. However, the conventional labeling period requires 15-18 h for robust biotinylation which may not be ideal for some applications. Recently, two new ligases termed TurboID and miniTurbo were developed using directed evolution of the BioID ligase and were able to produce robust biotinylation following a 10 min incubation with excess biotin. However, there is reported concern about inducibility of biotinylation, cellular toxicity, and ligase stability. To further investigate the practical applications of TurboID and ascertain strengths and weaknesses compared to BioID, we developed several stable cell lines expressing BioID and TurboID fusion proteins and analyzed them via immunoblot, immunofluorescence, and biotin-affinity purification-based proteomics. For TurboID we observed signs of protein instability, persistent biotinylation in the absence of exogenous biotin, and an increase in the practical labeling radius. However, TurboID enabled robust biotinylation in the endoplasmic reticulum lumen compared to BioID. Induction of biotinylation could be achieved by combining doxycycline-inducible expression with growth in biotin depleted culture media. These studies should help inform investigators utilizing BioID-based methods as to the appropriate ligase and experimental protocol for their particular needs.

Keywords: BioID; TurboID; biotinylation; lamin; nuclear pore complex; proximity-labeling.

Figure 1

Addition of exogenous biotin increases biotinylation in HA-tagged BioID, TurboID, and miniTurbo expressing cells. (A) A549 cell lysates stably expressing HA-tagged BioID, TurboID, or miniTurbo were analyzed via WB in the absence and presence of exogenous biotin (18 h). Anti-HA was used to probe for expression of the ligase and streptavidin HRP was used to probe for biotinylation. Anti-tubulin was used as protein loading control. (B) Confocal images of A549 cells stably expressing HA-tagged BioID, TurboID, or miniTurbo with and without the addition of excess biotin for 18 h. Anti-HA was used to visualize ligase localization (red) and fluorescently conjugated streptavidin was used to visualize biotinylation (green). Hoechst dye was used to detect the DNA in the nucleus (blue in merge). Scale bar 10 µm.

Figure 2

TurboID-fusion proteins biotinylate without the addition of biotin at similar levels to BioID- fusion proteins with biotin supplementation. Whole-cell lysates of A549 cells stably expressing BioID- or TurboID-tagged LaA, Nup43, or Nup53 fusion proteins were analyzed via western blot in the presence (+) of 50 μM biotin supplementation for 18 h or absence (−) of biotin supplementation. Anti-HA was used to detect fusion protein levels and streptavidin-HRP was used to probe for biotinylation. Anti-tubulin was used as a protein loading control.

Figure 3

A549 cells stably expressing TurboID-LaA biotinylate without the addition of exogenous biotin. (A) Western blot analysis of HA-tagged BioID-LaA and TurboID-LaA fusion protein expression (anti-HA) and overall biotinylation (Streptavidin-HRP) without or after biotin supplementation for 10 min or 18 h. Anti-tubulin was used as a protein loading control. (B) Confocal images of A549 cells stably expressing HA-tagged BioID-LaA and TurboID-LaA. Anti-HA was used to visualize fusion protein localization (red). Fluorescently conjugated streptavidin was used to detect biotinylation following no biotin supplementation or addition of biotin for 10 min or 18 h. Hoechst dye was used to detect the DNA in the nucleus (blue in merge). Scale bar 10 µm.

Figure 4

Post-pulldown proteomic analysis of LaA PPI candidates. (A) Following MS data analysis, BioID- and TurboID-LaA candidates were submitted to UniProt using the “Retrieve/ID Mapping” tool and subsequently grouped by subcellular domain designations. (B) Candidates were also submitted for STRING analysis visualization.

Figure 5

Nup43-TurboID-HA and HA-TurboID-Nup53 biotinylate without the addition of biotin. (A) Western blot analysis of A549 cell lysates expressing BioID- or TurboID-fused Nup43 or Nup53 in the absence or presence (10 min or 18 h) of biotin supplementation. Anti HA was used to detect fusion protein expression and streptavidin HRP was used to detect biotinylation. Anti-tubulin was used as a protein loading control. (B,C) Confocal visualization of BioID- or TurboID- Nup43 and Nup53. Fluorescently conjugated streptavidin was used to visualize biotinylation (green) following the addition (10 min or 18 h) or absence of 50 µM biotin. Anti-HA was used to visualize fusion protein expression and localization (red). Hoechst dye was used to detect the DNA in the nucleus (blue in merge). Scale bar 10 µm.

Figure 6

Schematics of practical labeling radii comparisons for Nup43 and Nup53 fusion proteins. (A) Schematic of labeling radii of Nup43-BioID and Nup43-TurboID within the Nup107-160 complex. Intensity of red labeling correlates with normalized percent total detected for each identified protein. (B) Schematic of labeling radii of Nup43-BioID and Nup43-TurboID within the NPC. Intensity of blue labeling correlates with normalized percent total detected for each identified protein. (C) Similar schematic as (B) for BioID-Nup53 and TurboID-Nup53.

Figure 7

Major candidates for TurboID-Nup53 are biotinylated prior to biotin supplementation. Protein IDs for each ligase and timepoint were submitted to the STRING database for cluster visualization.

Figure 8

Transient transfection of TurboID in A549 cells does not improve inducibility of biotinylation. (A) A549 cells transiently transfected with HA-TurboID-NLS were analyzed for expression (anti-HA) and biotinylation (Streptavidin-HRP) in the presence and absence of biotin (10 min) via WB. 24 h after transient transfection, cells were exposed to biotin for 10 min. Anti-tubulin was used as a protein loading control. (B) Epifluorescence images of A549 cells transiently transfected with HA-TurboID-NLS. Anti-HA was used to detect protein localization and fluorescently conjugated streptavidin was used to visualize biotinylation. Hoechst dye was used to detect the DNA in the nucleus (blue in merge). Scale bar 10 µm.

Figure 9

Inducible biotinylation can be achieved by combining dox-inducible TurboID expression with growth in biotin-depleted media. A549 cells stably expressing dox inducible HA-tagged TurboID were incubated with doxycycline and DMEM with 10% dialyzed serum for 18 h prior to the addition of biotin. The cells were then incubated with biotin for either 10 min, 1 h, or 4 h. Streptavidin-HRP was used to visualize biotinylation levels. Anti-HA was used to observe fusion protein levels. Anti-tubulin was used as a protein loading control.

Figure 10

TurboID is relatively stable within the ER lumen and biotinylates more efficiently than BioID. (A) Western blot analysis of A549 cells stably expressing HA-TurboID or HA-TurboID-Sun2 in the absence or presence (10 min, 2 h, or 18 h). A549 parental cells incubated with biotin for 18 h was used as a control for basal biotinylation levels. (B) Sun2-BioID-HA showed no biotinylation following 18 h of exogenous biotinylation, while Sun2-TurboID-HA showed biotinylation in the absence and presence (10 min, 2 h, 18 h) of incubation with excess biotin. Anti-HA (red) was used to visualize fusion protein localization and fluorescently conjugated streptavidin (green) was used to visualize biotinylation. Hoechst dye was used to detect the DNA in the nucleus (blue in merge). Scale bar 10 µm.