Fig 2: Activation of STAT3 promotes increased DNA binding and results in acute induction of STAT3 target genes. a, b Transcriptional activity of STAT3 is increased upon induction of DUBs and PDGF-BB stimulation. USP17L22 (a) or USP4 (b) were induced in the BJhTERT tet-inducible cell lines, cells were starved overnight and stimulated with PDGF-BB for 0, 15, 30 and 60 min. 20 µg of nuclear extracts were incubated with STAT3 binding element 5ʹ-TTCCCGGAA-3ʹ in a microplate and binding was quantified by colorimetric ELISA. Positive control was 5 µg of HepG2 cells stimulated with 100 ng/ml IL-6 (as supplied in the kit). The value for negative control was obtained by incubating each of the unstimulated or stimulated for 15, 30 or 60 min nuclear extracts from the uninduced BJhTERT cells with the addition of 20 pmol of unbound oligonucleotide that competed with the binding of activated STAT3 to the plate (as supplied in the kit). Absorbance values were plotted for three independent repeats; standard deviation is shown. The statistical significance was determined by one-tailed t test with unequal variance; *p < 0.05. c–h Induction of USP17L22 promotes acute upregulation of STAT3 target genes. USP17L22 was induced in the BJhTERT tet-inducible cell line which were stimulated with PDGF-BB for 1, 2 or 3 h and the expression of mRNA for STAT3 (c), CSF-1 (d), myc (e), SOCS3 (f), junB (g) and CDKN1A (h) was determined by quantitative PCR

Fig 3: Induction of USP17 and USP4 upregulate gene expression of myc and CDKN1A affecting cell proliferation. a–c USP17L22 promotes upregulation of myc and CDKN1A transcription in response to PDGF-BB. USP17L22 was induced in the BJhTERT tet-inducible cell line (a) and the expression of mRNA for myc (b) and CDKN1A (c) was determined by quantitative PCR. d–f USP4 affects expression of myc but not CDNK1A in response to PDGF-BB. USP4 was induced in BJhTERT tet-inducible cell line (d) and mRNA for myc (e) and CDKN1A (f) was determined by quantitative PCR. g Deletion of USP4 in BJhTERT CRISPR-Cas9 knockout cells decreases proliferative response to PDGF-BB, as compared to BJhTERT control cells. Fibroblasts were cultured for 3 days in the absence or the presence of increasing concentrations of PDGF-BB, cell proliferation was determined by measurement of DNA-intercalating fluorescent dye using CYQUANT proliferation assay. Relative fluorescent counts relate to the number of cells in the samples. h Total cell growth was not affected by deletion of USP4 in BJhTERT CRISPR-Cas9 knockout cells, while transient knockdown of USP17 in BJhTERT led to cell death. CYQUANT proliferation assay was used to measure relative number of cells growing under conditions of starvation in 1% FBS versus complete media for BJhTERT fibroblasts, BJ-CRISP-Cas9-USP4 fibroblasts (BJ-KO-Usp4) and BJhTERT fibroblasts transiently transfected with nontargeting control siRNA or USP17L9P siRNA

Fig 4: USP4 and USP17L22 affect the timing of activation of STAT3. a, b Induction of USP17L22 results in a shift of STAT3 activation to later time points in BJhTERT inducible cells. Expression of Flag-tagged USP17L22 was induced with doxycycline, cells were serum-starved and stimulated with PDGF-BB (20 ng/ml) for the indicated time periods. Expression of total and phosphorylated proteins was determined in total cell lysates using antibodies against PDGFRß, Flag, STAT3 (pY705), STAT3, PLC? (pY783), PLC?, pS473 Akt1/2/3, Akt1/2/3, pThr202/pThr204 Erk1/2, Erk1/2 and ß-actin (a). Phosphorylated STAT3 (pSTAT3) versus total STAT3 levels were quantified in three experiments; peak of phosphorylation at 10 min of stimulation with PDGF-BB was set as 1. Standard deviation is shown between repeats; *p < 0.05 (b). c, d Induction of USP4 increases the STAT3 activation in BJhTERT-USP4 inducible cells. BJhTERT-USP4 inducible cells were treated as described and immunoblotting was performed for proteins as in panel a; a-tubulin was used as loading control. Activated pSTAT3 relative to STAT3 protein levels were quantified in three independent experiments as in panel b; *p < 0.05 (d). e CRISPR-Cas9 knockout of USP4 decreases the STAT3 activation in BJhTERT-CRISPR-Cas9-USP4 knockout cells. The control BJhTERT and CRISPR-Cas9-USP4 knockout BJhTERT cells were starved overnight and stimulated with PDGF-BB (20 ng/ml) for the indicated time periods. Immunoblotting was performed for the indicated proteins, as in panels a–d. f USP4 does not deubiquitinate STAT3. After USP4 induction and serum starvation, BJhTERT-USP4 cells were stimulated with PDGF-BB (20 ng/ml) for the indicated time periods, lysates were divided and immunoprecipitated with anti-PDGFRß (CTß) or anti-STAT3 antibodies. Eluates were immunoblotted for ubiquitin. Levels of PDGFRß, Flag-USP4 and a-tubulin were determined by immunoblotting (IB)

Fig 5: USP4 regulates the speed of co-localization of PDGFRβ with early endosomes. a, b Induction of USP4 delays co-localization of PDGFRβ with early endosomes. USP4 expression was induced in BJhTERT-USP4 tet-inducible cell line versus noninduced control cells. Representative images of co-localization (in yellow) of PDGFRβ (red) with early endosomal marker EEA1 (green) at indicated times of stimulation with PDGF-BB (a). Co-localization of PDGFRβ and EEA1 was quantified in 3 independent experiments and is presented as Pearson coefficient of correlation of distribution of signals for each analysed channel in 15–35 cells per condition. *p < 0.05 (b). c–e CRISPR-Cas9-mediated knockout of USP4 leads to faster co-localization of PDGFRβ with early endosomes. Co-localization of PDGFRβ and EEA1 marker was quantified and presented as in panel b for BJ-CRISPR-Cas9 knockout cells versus BJhTERT control cells (c) and U2OS CRISP-USP4 knockout cells versus U2OS control cells (d). e Representative images of co-localization (in yellow) of PDGFRβ (red) with early endosomal marker (green) at indicated times of stimulation with PDGF-BB of the experiments quantified in panel c are presented. f Co-localization between PDGFRβ and late endosomal marker (Rab7) is presented for BJhTERT CRISP-USP4 knockout cells versus BJhTERT control cells