PROTAC-D treatment leads to greater degradation of nuclear AURKA compared to cytoplasmic AURKA

A Representative images of U2OS FZR1KO cells transfected with pNeonN1-AURKA (wildtype, WT) or nuclear-localised pNeonN1-NLS-AURKA (NLS) and treated with DMSO or PROTAC-D (1 µM) for 4 hours. Images acquired by widefield fluorescence microscopy. Scale bar 10 µm.

B Quantification of the nuclear-to-cytoplasmic fluorescence intensity ratio for WT- and NLS- AURKA-Neon constructs shown in (A). Data show pooled single-cell measurements with bars representing mean ± SD from one experiment (n ≥ 13 cells) and are representative of three biological repeats; statistical analysis by unpaired t-test.

C DMSO-normalised mNeon fluorescence of WT- or NLS-AURKA-Neon in U2OS FZR1KO cells after 4-hour treatment with DMSO or PROTAC-D (1 µM). Images were acquired using time- lapse microscopy and mNeon fluorescence levels in single cells measured at 0- and 4-hour timepoints. Substrate degradation is expressed as percentage change of fluorescence at 4 hours in single cells, normalised against the mean value from DMSO-treated cells and data points are mean values ± SD from three biological repeats (≥ 18 cells from multiple fields analysed per condition across all repeats). Statistical significance determined by two-way ANOVA with Tukey’s post hoc multiple comparison test.

D U2OS FZR1KO cells co-transfected with pNeonN1-AURKA/pNeonN1-NLS-AURKA and pcDNA3-FLAG-Ub were treated with DMSO or PROTAC-D (100 nM) for one hour. Ubiquitinated proteins from cell extracts were captured by pulldown with ubiquitin affinity beads. The band marked (*) is a non-specific band.

E Quantification of PROTAC-D-induced ubiquitin conjugates shown in (D), expressed as the ratio of ubiquitin conjugates to input detected via the HA tag on mNeon, normalised to the ratio for WT-AURKA-Neon. The ubiquitin smear was quantified above the non-specific band at around 110 kDa, marked (*).

F Representative images of mNeon/CRBN isPLA signal, indicative of ternary complex formation, detected in interphase U2OS FZR1KO cells transfected with pNeonN1-AURKA. Cells were treated for 1 hour with DMSO or PROTAC-D (1 µM) and processed for isPLA using mNeon and CRBN primary antibodies in combination or alone to control for non-specific signal. Scale bar 10 µm.

G Quantification of isPLA signal from (F). Total isPLA intensity was measured in single cells and pooled data plotted as mean ± SEM (n ≥ 13 cells). Statistical significance was assessed using one-way ANOVA with Tukey’s post hoc multiple comparisons test.

H Representative images of mNeon/CRBN isPLA signal in interphase U2OS FZR1KO cells transfected with pNeonN1-AURKA or pNeonN1-NLS-AURKA after 1 hour treatment with DMSO or PROTAC-D (1 µM). Scale bar 10 µm.

I Quantification of isPLA signal from (H). Total isPLA intensity was measured in single cells and pooled data plotted as mean ± SEM (n ≥ 19 cells). Statistical significance was assessed using two-way ANOVA with Tukey’s post hoc multiple comparisons test.

siRNA screen identifies DUBs regulating HiBiT-AURKA levels and sensitivity to PROTAC-D-mediated degradation.

A HiBiT-AURKA bioluminescent signal normalised to DMSO controls in doxycycline-inducible U2OS HiBiT-AURKATO cells after 4-hour treatment with DMSO, PROTAC-D, PROTAC-DX or PROTAC-D plus proteasome inhibitor MG132, at the concentrations indicated. Individual data points represent mean values from three technical replicates per experiment, with bars showing mean value ± SD from three independent experiments (n=3). Statistical analysis by one-way ANOVA with Dunnett’s post-hoc multiple comparison test to DMSO.

B Dose-response curve of DMSO-normalised HiBiT-AURKA bioluminescence, showing means ± SD from n=3 biological repeats.

C Dose-response curve for the CellTiter-Glo®-based cell viability assay normalised to DMSO. Data represent means ± SD from n=3 biological repeats.

D Workflow of the siRNA screen used to identify DUBs regulating HiBiT-AURKA levels. Created with BioRender.com.

E Percentage change in PROTAC-D-mediated HiBiT-AURKA degradation over 4 hours, relative to negative control siRNA. Average values from two biological repeats are plotted.

F Percentage change in HiBiT-AURKA levels in DMSO-treated cells relative to negative control siRNA. Average values from two biological repeats are plotted.

G Summary of top-scoring DUB hits for those affecting PROTAC-D-mediated degradation (left) and those affecting basal HiBiT-AURKA levels (right).

validation of DUB hits from siRNA screen as PROTAC-opposing DUBs

A U2OS HiBiT-AURKATO cells transfected with siRNA pools (4 oligos) targeting the indicated DUBs were treated with DMSO or PROTAC-D (100 nM) for 4 hours. HiBiT-AURKA bioluminescence was quantified, and PROTAC-D-mediated degradation was normalised to DMSO controls. Data are plotted as the percent change in degradation relative to negative control siRNA, with means ± SD from n=3 biological repeats.

B, C U2OSCDK2 cells were transfected with siRNA pools targeting the indicated DUBs. 48 hours post-transfection nuclear-to-cytoplasmic DHB-Cer signal ratios were measured from widefield fluorescence images of live cells. B Data shows median DHB-Cer cytoplasmic:nuclear ratio plotted against percentage change in PROTAC-D activity as determined in (A), for each candidate DUB siRNA. Medians are the average values from two biological repeats, percentage changes in PROTAC-D activity are the average values from three biological repeats. Simple linear regression is plotted with Pearson’s statistical test for correlation. C Bars show the proportion of cells in each DUB siRNA-transfected pool with a cytoplasmic:nuclear ratio < 1, plotted as mean values ± SD from n=2 biological repeats, with ≥36 cells analysed per experiment.

D Representative images of U2OS cells transfected with negative control siRNA and treated with DMSO or PROTAC-D (100 nM), stained for AURKA (green) and tubulin (red). DNA stained with DAPI. Scale bar 10 µm.

E Quantification of whole-cell AURKA levels in pre-anaphase mitotic U2OS cells such as the examples shown in (D), transfected for 48 hours with indicated siRNA and treated with DMSO (upper panels) or PROTAC-D (lower panels) for 4 hours. AURKA intensity was measured in single cells and normalised to tubulin intensity. Data clouds show values from individual cells, with bars indicating mean values ± SEM for ≥ 15 cells analysed per condition. Statistical significance assessed via t-test.

UCHL5 opposes PROTAC-mediated degradation of AURKA and other substrates

A DMSO-normalised HiBiT-AURKA levels after 4-hour treatments with DMSO, PROTAC-D, JB170, or TL12-186 (100 nM) in U2OS HiBiT-AURKATO cells. Data represent means ± SD from n=3 biological repeats; statistical analysis by one-way ANOVA with Dunnett’s post-hoc multiple comparisons test to DMSO-treated cells.

B Immunoblot analysis of UCHL5 levels in U2OS HiBiT-AURKATO cells transfected with negative control siRNA (GL2i) or UCHL5-targeting siRNAs (two oligos). Lysates were collected 48 hours post-transfection.

C U2OS HiBiT-AURKATO cells transfected with GL2 siRNA or UCHL5 siRNA were treated with DMSO, PROTAC-D, JB170 or TL12-186 (100 nM). HiBiT-AURKA bioluminescence signal was normalised to DMSO-treated controls for each transfection condition. Percentage change in PROTAC-mediated HiBiT-AURKA degradation relative to negative control siRNA is plotted, with individual data points corresponding to mean values of technical replicates from n=3 biological repeats and line to indicate the mean value of biological repeats; statistical analysis by unpaired t-test.

D siRNA-rescue experiment in U2OS HiBiT-AURKATO cells transfected with combinations of GL2 siRNA, UCHL5 siRNA, and siRNA-resistant pNeonN1-UCHL5 (siR). Cells were treated with DMSO, PROTAC-D, or TL12-186 (100 nM) for 4 hours, and HiBiT-AURKA signal was normalised to DMSO-treated controls. Percentage change in PROTAC-mediated HiBiT-AURKA degradation relative to negative control siRNA is plotted. Data points correspond to mean values of technical replicates from n=3 biological repeats with line to indicate the mean value of biological repeats; statistical analysis by one-way ANOVA with Tukey’s post hoc multiple comparisons test.

E Dose-response curves of U2OS cells transfected with pNeonN1-HiBiT-FKBP12F36V or pcDNA3.1-HiBiT-AURKA and treated with varying doses of dTAG-13 (upper panel) or dTAGv-1 (lower panel). Data are DMSO-normalised means ± SD from n=3 biological repeats.

F, G U2OS cells transfected with pNeonN1-HiBiT-FKBP12F36V and GL2 siRNA or UCHL5 siRNA were treated with DMSO, dTAG-13 (F), or dTAGv-1 (G) at the indicated concentrations for 4 hours. HiBiT bioluminescent signal was normalised to DMSO-treated controls. Percentage change in PROTAC-mediated HiBiT-FKBP12F36V degradation in UCHL5 siRNA relative to negative control siRNA is plotted. Data points correspond to mean values of technical replicates from n=3 biological repeats with line to indicate the mean value of biological repeats; statistical analysis by unpaired t-test.

OTUD6A opposes PROTAC-mediated degradation of AURKA in a target-specific manner

A Immunoblot analysis of U2OS HiBiT-AURKATO cells transfected with negative control siRNA (GL2i) or OTUD6A-targeting siRNAs (two oligos). Lysates were collected 48 hours post- transfection.

B U2OS HiBiT-AURKATO cells transfected with GL2 siRNA or OTUD6A siRNA were treated with DMSO, PROTAC-D, JB170 or TL12-186 (100 nM). HiBiT-AURKA signal was normalised to DMSO- treated controls for each transfection condition. Data points show percentage change in PROTAC-mediated HiBiT-AURKA degradation relative to negative control siRNA and correspond to mean values of technical replicates from n=3 biological repeats, with line to indicate the mean value of biological repeats; statistical analysis by one-way ANOVA with Tukey’s post hoc multiple comparisons test.

C siRNA-rescue experiment in U2OS HiBiT-AURKATO cells transfected with combinations of GL2 siRNA, OTUD6A siRNA, and siRNA-resistant pNeonN1-OTUD6A (siR). Cells were treated with DMSO, PROTAC-D, or TL12-186 (100 nM) for 4 hours, and HiBiT-AURKA signal was normalised to DMSO-treated controls. Data points show percentage change in PROTAC-mediated HiBiT- AURKA degradation relative to negative control siRNA and correspond to mean values of technical replicates from n=3 biological repeats, with line to indicate the mean value of biological repeats; statistical analysis by one-way ANOVA with Tukey’s post hoc multiple comparisons test.

D U2OS HiBiT-AURKATO cells transfected with pEGFP, pEGFP-OTUD6A, or catalytically inactive pEGFP-OTUD6AC152A plasmids were treated with DMSO or PROTAC-D (100 nM) for 4 hours and HiBiT-AURKA signal was normalised to DMSO-treated condition. Data points show percentage change in PROTAC-mediated HiBiT-AURKA degradation relative to pEGFP overexpression and correspond to mean values of technical replicates from n=3 biological repeats, with line to indicate the mean value of biological repeats; statistical analysis by one-way ANOVA with Tukey’s post hoc multiple comparisons test.

E, F U2OS cells transfected with pNeonN1-HiBiT-FKBP12F36V and GL2 or OTUD6A siRNA were treated with DMSO, dTAG-13 (E), or dTAGv-1 (F) at the indicated concentrations for 4 hours. Data points show percentage change in PROTAC-mediated HiBiT-FKBP12F36V degradation relative to negative control siRNA, plotted as mean values of technical replicates from n=3 biological repeats, with line to indicate the mean value of biological repeats; statistical analysis by unpaired t-test.

G U2OS FZR1KO cells co-transfected with pNeonN1-AURKA-Neon-HA, pcDNA3-FLAG-Ub and GL2, OTUD6A or UCHL5 siRNA were treated with DMSO or PROTAC-D

(100 nM) for one hour. Ubiquitinated proteins from cell extracts were captured by pulldown with ubiquitin affinity beads.

H Quantification of (G): chart shows the ratio of ubiquitin conjugates to input after PROTAC-D treatment for each condition, normalised to the ratio for GL2i. The quantified ubiquitin smear region was above the non-specific band at around 110 kDa, marked (*).

Enhanced degradation of the nuclear pool of AURKA in response to PROTACs is substrate-specific

A Schematic of FKBP12F36V-Neon constructs and engagement with CRBN-recruiting dTAG-13 and VHL-recruiting dTAGv-1 PROTACs. Created with BioRender.com.

B Representative fluorescence images of U2OS FZR1KO cells transfected with pNeonN1- AURKA, pNeonN1-AURKA-FKBP12F36V or pNeonN1-FKBP12F36V, treated with PROTAC-D (1 µM) or dTAG-13 (500 nM) for 4 hours. Scale bar 10 µm.

C Quantification of experiments illustrated in (B). Images were acquired using time-lapse microscopy and mNeon fluorescence levels in single cells measured at 0- and 4-hour timepoints. Substrate degradation is expressed as percentage change of fluorescence at 4 hours in single cells, normalised against the mean value from DMSO-treated cells. Bar chart shows mean values ± SD from three biological repeats, with ≥ 15 cells from multiple fields analysed per condition across all repeats. Statistical significance determined by two-way ANOVA with Tukey’s post hoc multiple comparison test.

D, E U2OS FZR1KO cells were transfected with WT- or NLS-AURKA-FKBP12F36V-Neon and subjected to timelapse fluorescence imaging after treatment with DMSO, dTAG-13 (D) or dTAGv-1 (E) at the indicated concentrations. mNeon fluorescence was quantified in single cells at 0- and 4-hour timepoints and change in fluorescence at 4 hours normalised for the mean values from DMSO-treated cells. Bar chart shows mean values ± SD from three biological repeats, with ≥ 20 cells from multiple fields analysed per condition across all repeats. Statistical significance determined by two-way ANOVA with Tukey’s post hoc multiple comparison test.

F, G U2OS FZR1KO cells were transfected with WT- or NLS-FKBP12F36V-Neon and subjected to timelapse fluorescence imaging after treatment with DMSO, dTAG-13 (F) or dTAGv-1 (G) at the indicated concentrations. mNeon fluorescence was quantified in single cells at 0- and 4-hour timepoints and change in fluorescence at 4 hours normalised for the mean values from DMSO- treated cells. Bar chart shows mean values ± SD from three biological repeats, with ≥ 22 cells from multiple fields analysed per condition across all repeats. Statistical significance determined by two-way ANOVA with Tukey’s post hoc multiple comparison test.

Enhanced degradation of the nuclear pool of AURKA is OTUD6A-dependent

A U2OS cells transfected with pEGFP-OTUD6A, showing subcellular localisation of GFP- OTUD6A in interphase cells. Scale bar 10 µm.

B U2OS FZR1KO cells were transfected with WT- or NLS-AURKA-Neon and GL2 siRNA or OTUD6A siRNA for 48 hours then treated for 4 hours with DMSO or PROTAC-D (1 µM) and subjected to timelapse fluorescence imaging. mNeon fluorescence was quantified in single cells at 0- and 4-hours and change in fluorescence normalised to the mean value from DMSO-treated cells. Bar charts show mean values ± SD from three biological repeats, with ≥ 17 cells from multiple fields analysed per condition across all repeats. Statistical significance determined by two-way ANOVA with Tukey’s post hoc multiple comparison test.

C U2OS FZR1KO cells were transfected with WT- or NLS-AURKA-Neon, with or without pN1- mCherry-OTUD6A for 24 hours then treated for 4 hours with DMSO or PROTAC-D (1 µM) and subjected to timelapse fluorescence imaging. mNeon fluorescence was quantified in single cells at 0- and 4-hours and change in fluorescence normalised to the mean value from DMSO- treated cells. Image panels show transfected cells at timepoint 0. Bar chart shows mean values ± SD from three biological repeats, with ≥ 14 cells from multiple fields analysed per condition across all repeats. Statistical significance determined by two-way ANOVA with Tukey’s post hoc multiple comparison test.

PROTAC-induced ubiquitination of AURKA-Neon-HA

U2OS FZR1KO cells co-transfected with pNeonN1-AURKA-HA and pcDNA3 FLAG-Ub were treated as indicated for one hour. Ubiquitinated proteins from cell extracts were captured by pulldown with ubiquitin affinity beads. The band marked (*) is a non-specific band. Treatment with p97 inhibitor (CB-5083), but not proteasome inhibitor (MG132), enriched pulldowns for polyubiquitinated AURKA.

AURKB is sensitive to PROTAC-DX

A Immunoblot of U2OS cells treated with DMSO, PROTAC-DX or PROTAC-D (1 µM) for 4 hours prior to lysate collection. Panels are from the same immunoblot.

B Quantification of vinculin-normalised AURKA and AURKB levels from the immunoblot in (A). Bar chart shows mean values ± SD from two or three biological repeats.

Cell cycle influence on PROTAC-D activity

A Schematic representation of the CDK2 activity sensor expressed in U2OSCDK2 cells. Created with BioRender.com.

B U2OS HiBiT-AURKATO cells were synchronised as indicated and treated for 4 hours with DMSO or PROTAC-D (100 nM). Bar chart shows DMSO-normalised HiBiT-AURKA levels as mean values ± SD from n=3 biological repeats; statistical analysis by one-way ANOVA with Dunnett’s post hoc multiple comparisons to unsynchronised cells.

PROTAC-induced ubiquitination of HiBiT-FKBP12F36V-Neon-HA

A U2OS HiBiT-AURKATO cells transfected with pEGFP, pEGFP-UCHL5, or catalytically inactive pEGFP-UCHL5C88A plasmids were treated with DMSO or PROTAC-D (100 nM) for 4 hours. Data points show percentage change in PROTAC-mediated HiBiT-AURKA degradation relative to pEGFP overexpression and correspond to mean values of technical replicates from n=3 biological repeats with line to indicate the mean value of biological repeats; statistical analysis by one-way ANOVA with Tukey’s post hoc multiple comparisons test.

B U2OS cells co-transfected with pNeonN1-HiBiT-FKBP12F36V and pcDNA3-FLAG-Ub were treated with DMSO, dTAG-13 or dTAGv-1 (100 nM) and CB-5083 (5 µM) for one hour. Ubiquitinated proteins from cell extracts were captured by pulldown with ubiquitin affinity beads.

C Quantification of (B): chart shows the ratio of ubiquitin conjugates to input for each condition, normalised to the ratio for DMSO treatment.