Figure 1.
Design of ubiquitination-induced fluorescence complementation (UiFC) assay.
(A) Schematic representation of UiFC-N and UiFC-C and the proposed mechanism of UiFC for detecting polyubiquitin chain. (B) UiFC-N and UiFC-C interact with K11, K48, and K63 chains. Purified 6His-UiFC-N and 6His-UiFC-C were incubated with K11, K48, or K63 chains. Following the incubation immunoprecipitation was carried out with anti-ubiquitin antibody followed by immunoblotting for venus. Control IP: anti-HA antibody was used instead of anti-ubiquitin. Input: 10% of total UiFC-N and UiFC-C or polyubiquitin chains. Aristerisk indicates UiFC-N. Arrowhead indicates UiFC-C. (C) Spontaneous reconstitution of the N and C-terminal fragment of venus in UiFC-N and UiFC-C in vitro. Increasing amounts of 6His-UiFC-N and 6His-UiFC-C were incubated in 20 µl Tris buffer and venus fluorescence was measured every 10 minutes for up to 270 minutes at 37 °C as described in Experimental Procedures. The mean +/- SD of venus fluorescence intensities from triplicate reactions at each time point was plotted. (D) K48 ubiquitin chain but not mono-ubiquitin induces increases in venus fluorescence. UiFC-N and UiFC-C (0.86 µM) were incubated with the reaction mix for K48 chain synthesis containing 2 µM ubiquitin in the presence or absence of ATP for 60 minutes. Control: no K48 chains, DUB: 6His-EBV-DUB. The graph shows the mean +/- SD of UiFC fluorescence intensities from triplicate reactions (left). (E) The reaction mixes used in D were analyzed by immunoblotting for ubiquitin after the measurement of venus fluorescence. Lower panel: the added 6His-EBV-DUB revealed by Ponceau staining.
Figure 2.
UiFC-N and UiFC-C preferentially detects with K48 ubiquitin chains in vitro.
(A) Characterization of polyubiquitin chains with K11, K48, or K63 linkage. Ubiquitin chains were assembled in vitro (see Experimental Procedures for details) and analyzed by immunoblotting with linkage-specific antibody. The blots were from the same membrane stripped and re-blotted with antibody against total ubiquitin, K11 chain, K48 chain, and K63 chain, respectively. (B) Increasing amounts of K11, K48, or K63 chains (formed by18, 36, and 72 pmols of mono-ubiquitin) and mono-ubiquitin (72 pmols) were analyzed by immunoblotting for ubiquitin. Mono-ubiquitin was added to make the total amount of ubiquitin equal to 72 pmols in all lanes. (C) Measurement of UiFC fluorescence in reactions containing increasing amounts of ubiquitin chains, 6His-UiFC-N, and 6-His-UiFC-C at 10 minutes interval for up to 210 minutes. Mono-ubiquitin was added to make the final concentration of ubiquitin (mono or in chains) equal to 3.6 µM in all reactions. The mean +/- SD of the UiFC fluorescence intensities from triplicate reactions at each time point was plotted.
Figure 3.
Characterization of UiFC assay in vitro.
(A) Generation of UiFC fluorescence is time and K48 chain concentration-dependent. Increasing concentrations of K48 chains were incubated with 6His-UiFC-N and 6His-UiFC-C and the UiFC fluorescence was measured during incubation at 10 minutes interval for 280 minutes. The mean +/- SD of the UiFC fluorescence intensities from triplicate reactions at each time point was plotted.
The lines from bottom to top represent 0.45, 0.9, 1.8, 3.6, 5, and 6 µM ubiquitin chains, respectively. (B) The maximal UiFC fluorescence intensity correlates well with K48 chain concentration but time required to achieve maximal fluorescence is much less dependent on ubiquitin chain concentration. (C) Kinetics of UiFC fluorescence in the detection of K48 chain. Time-dependent increases in fluorescence (ΔF/Δt) in linear region shown in A are plotted against K48 chain concentration. (D) Z factor of UiFC. 15 UiFC assays were performed and Z factor is calculated as described in Experimental Procedure. An example is shown in the graph.
Figure 4.
UiFC fluorescence colocalize well with K48 chains in cells.
(A) Expression of UiFC-N alone does not produce any venus fluorescence in cells. HeLa cells were transfected with plasmid encoding UiFC-N and stained with anti-ubiquitin antibody (FK2) (left panel). (B) Co-localization of UiFC fluorescence (yellow) with anti-K48 chain immunofluorescence (blue, pseudo-colored to cyan to increase visibility) except for the large intranuclear UiFC aggregates (asterisk, see text for explanation). UiFC-N and UiFC-C co-transfected cells were stained with anti-K48 chain antibody. Inset: co-localization of UiFC puncta and K48 chains. (C) UiFC fluorescence (yellow) and K63 chains (red) are largely localized in different subcellular structures. Plasmids encoding UiFC-N, UiFC-C, and mCherry-tagged K63 chain sensor were transfected into HeLa cells. (D) Inhibition of proteasome but not autophagy pathway increases UiFC fluorescence. Co-expression of UiFC-N and UiFC-C produces diffused fluorescence and fluorescent puncta (yellow) throughout the cell (left panel); Treatment with MG132 (10 µM, middle panel) but not 3-MA (20 mM, right panel) for 4 hours leads to increases in UiFC fluorescence. UiFC: UiFC-N and UiFC-C. (E) Quantification of UiFC fluorescence intensities in control cells and cells treated with MG132 or 3MA. The fluorescence intensity of individual cell was measured using ImageJ software. The graph shows the mean +/- SD of UiFC fluorescence, n ≥ 53. (F) Effects of proteasome inhibition on UiFC-N and UiFC-C levels. HeLa cells co-transfected with UiFC-N and UiFC-C were treated with MG132 (10 µM) for the indicated time and the levels of UiFC-N and UiFC-C were determined by immunoblotting. Polyubiquitin was blotted to show the efficiency of MG132. Actin was blotted as a loading control.
Figure 5.
UiFC fluorescence colocalize well with K48 chains but not K63 chains enriched in p62/SQSTM1 bodies.
K63 chains colocalize well with p62/SQSTM1 bodies (lower left) but not with UiFC fluorescence (lower right). HeLa cells transfected with plasmids encoding HA-p62/SQSTM1 and UiFC were processed for immunostaining with mouse monoclonal anti-p62/SQSTM1 (blue) and rabbit monoclonal anti-K63 chain antibody (red) (A) or affinity-purified rabbit polyclonal anti-p62/SQSTM1 (blue, pseudo-colored to cyan) and mouse monoclonal anti-K48 chain antibody (red) (B), followed by fluorescent microscopy.
Figure 6.
The mitochondrial outer membrane localization of UiFC in Parkin-expressing and FCCP-treated cells.
(A) Parkin-dependent UiFC fluorescence associated with the mitochondria. HeLa cells co-expressing UiFC-N, UiFC-C and CFP-Parkin were treated with the uncoupler FCCP to induce mitochondrial damage and trigger mitochondrial translocation of Parkin. FCCP induced UiFC fluorescence associated with mitochondria in two cells expressing CFP-Parkin. No perinuclear UiFC aggregates seen in the cell that does not express CFP-Parkin (*). (B, C) Colocalization of UiFC and Tom20 or substrate-conjugated ubiquitin. HeLa cells co-expressing UiFC-N, UiFC-C and untagged Parkin were treated with FCCP. Cells were then processed for immunostaining for the marker of the outer mitochondrial membrane Tom20 (B), or with FK2 antibody that recognize substrate-conjugated ubiquitin (C) and analyzed by fluorescence microscopy. Tom20 and FK2 signals are red; UiFC is yellow on overlay images. The cell expressing higher level (arrows in B) but not the lower level (*) of UiFC exhibits intranuclear UiFC aggregates. The colocalization of Tom20 or ubiquitin (red) and UiFC (yellow) was also shown by the profiles of the fluorescence along blue lines in the enlarged images.
Figure 7.
Application of UiFC to image Parkin-dependent ubiquitination of mitochondrial outer membrane proteins.
(A) Cells co-expressing UiFC-N, UiFC-C (yellow on overlay images) and mCherry-Parkin (red on overlay images) were treated with the uncoupler FCCP to induce mitochondrial damage and mitochondrial translocation of mCherry-Parkin. An example of the time-lapse studies is shown. In (A) the same cells at time 0 and 70 min after addition of FCCP are depicted. In (B) details from boxed areas in (A), as well as plot profiles along the yellow lines in overlay detail images are shown (red is mCherry-Parkin, yellow UiFC). (C) Detailed time-lapse analyzes of the changes in mCherry-Parkin and UiFC over the time of FCCP-treatment in cells shown in (A) are depicted. Changes in the mCherry-Parkin and the UiFC fluorescence intensities normalized to the cytosolic background fluorescence intensities were also quantified and plotted as a function of treatment time (D).
Figure 8.
UiFC in live cells under stress.
(A) Examples of time-lapse experiments of untreated (Ctr: control), proteasome inhibitor MG132 (MG, 10µM)-, or tunicamycin (Tm, 2µg/ml)-treated cells transfected with UiFC-N and UiFC-C. Cells were imaged every 10 min for up to 240 minutes. (B, C) Graphs show changes in the UiFC fluorescence intensities, normalized as a fold increase over the initial intensity (taken as 1 at time 0 min) in untreated cells or cells treated with MG132, tunicamycin, or doxorubicin. Inset in (B) shows changes in UiFC fluorescence every min for 9 min of MG132 treatment. Data in (B and C) represent mean +/-SD (n > 70 in all experimental groups). (D) Co-localization of K48 chains (blue) and UiFC fluorescence (yellow) under ER stress. HeLa cells co-expressing UiFC-N and UiFC-C were treated with tunicamycin treatment for 4 hours as in (A) followed by immunofluorescent staining for K48 chains (blue). (E) Effects of UiFC-N and UiFC-C on ubiquitination in vitro. Increasing amounts of 6His-UiFC-N and 6His-UiFC-C was added in the reaction mix contains human E1, Ube2g2, gp78C, ubiquitin and ATP (see details in Experimental Procedures) (ATP+) and incubated for 3 hours followed by measuring UiFC fluorescence on a fluorescent microplate reader. The same reaction mix without ATP (ATP-) or with ATP and EBV-DUB (DUB+) was performed as controls. The graph shows mean +/- SD of triplicate wells. The reaction mixes were analyzed by immunoblotting for ubiquitin after the measurement of UiFC fluorescence. (F) UiFC-N and UiFC-C concentration-dependent effects on ubiquitination in vitro. Increasing amount (0, 0.125, 0.25, 0.5, 1, 2, and 4 µg of each) of 6His-UiFC-N and 6His-UiFC-C were included in in vitro polyubiquitination mix as described in (E). Ubiquitin chains were analyzed by immunoblotting for ubiquitin.