A novel USP9X substrate TTK contributes to tumorigenesis in non-small-cell lung cancer.

The X-linked deubiquitinase, USP9X, is implicated in multiple cancers by targeting various substrates. Increased expression of USP9X is observed in non-small-cell lung cancer (NSCLC) and is correlated with poor prognosis. However, the molecular mechanism for USP9X regulation of tumor cell survival and tumorigenesis in NSCLC is less defined.
Methods: In this study, chemical labeling, quantitative proteomic screening was applied to analyze A549 cells with or without USP9X RNA interference. Functional in vitro and in vivo experiments were performed to confirm the oncogenic effects of USP9X in NSCLC and to investigate the underlying mechanisms.
Results: The resulting data suggested that dual specificity protein kinase TTK is a potential substrate of USP9X. Further experimental evidences confirmed that USP9X stabilized TTK via direct interaction and efficient deubiquitination of TTK on K48 ubiquitin chain. Moreover, knockdown of USP9X or TTK inhibited cell proliferation, migration and tumorigenesis, and the immunohistochemical analysis of clinical NSCLC samples showed that the protein expression levels of USP9X and TTK were significantly elevated and positively correlated in tumor tissues. Conclusions: In summary, our data demonstrated that the USP9X-TTK axis may play a critical role in NSCLC, and could be considered as a potential therapeutic target.

Introduction

Lung cancer is one of the leading causes of cancer death worldwide 1. Incidence and mortality attributed to lung cancer have risen steadily since the 1930s 2. If diagnosed early, survival of patients suffering from lung cancer could be improved primarily due to advances in treatments. NSCLC accounts for ~85% of all types of lung cancers, but is often diagnosed at the late or advanced stage and has poor prognosis 3. Therefore, further study is necessary to discover biomarkers and therapeutic targets for NSCLC prognosis and therapy.

Ubiquitination, as a major post-translational modification involved in protein activation, localization and degradation, participates in various biological processes 4. Protein ubiquitination is a reversible process catalyzed by ubiquitin ligases and deubiquitinases (DUBs). DUBs catalyze the release of ubiquitin from substrate proteins and oppose the action of ubiquitin ligases. In humans, DUBs are subdivided into five families: USPs, OTUs, UCHs, Josephins and JAMMs 5. DUBs have been extensively studied due to their important roles in diverse biological processes including DNA repair 6, chromatin remodeling 7, cell-cycle control 8, embryonic stem cells stabilization 9 and particular functions in tumorigenesis 10-14. DUBs regulating processes associated with cell proliferation and apoptosis are considered candidate targets for cancer therapeutics 15.

Among ~90 DUBs, the X-linked deubiquitinase, USP9X, regulates the components of multiple signaling pathways and can act as an oncogene or a tumor suppressor in different cancers. For example, USP9X is considered a tumor suppressor in pancreatic ductal adenocarcinoma 16, 17. In contrast, in melanoma 18, esophageal squamous cell carcinoma 16, 19, and lung cancer 20, 21, overexpression of USP9X correlates with poor prognosis, which indicates its oncogenic role. Therefore, in-depth study of USP9X is important to demonstrate the mechanism of USP9X involvement in tumorigenesis and further cancer therapy.

Dual specificity protein kinase TTK, also known as Mps1, plays crucial roles in spindle pole assembly, spindle checkpoint, genotoxic stress response, cytokinesis and meiosis 22. The expression level of TTK is tightly regulated during cell-cycle progression 22. TTK mRNA levels are elevated in breast cancer 23, lung cancer 24 and thyroid carcinoma 25. TTK is degraded by the ubiquitin-proteasome pathway 26. The ubiquitin ligases anaphase-promoting complex/ cyclosome (APC-c), including APC-cCdc20 and APC-cCdh1, have been indicated to target TTK in a cell cycle-dependent manner 26. And, another previous study also found that ubiquitin ligase Ufd2 is required for efficient degradation of TTK 27. However, little is known about the deubiquitinases of TTK.

The protein interaction partners of DUBs and their substrates can be identified by mass spectrometry-based methods. The DUBs binding ubiquitinated substrates complexes can be enriched and identified using tagging affinity-based approaches followed by tandem mass spectrometry analysis. For example, this approach was successfully applied to identify CEP131 28 as a substrate of USP9X. Another useful strategy for discovering the substrates of DUBs is to investigate the changed proteins in the cells with knockdown/overexpression of the DUBs by global proteomics. Poulsen et al. applied the SILAC-based quantitative proteomic approach to compare wild-type Saccharomyces cerevisiae to 20 DUBs knock-out strains, and demonstrated that Sec28p is a novel substrate of Ubp3p 29. In this study, we utilized a chemical labelling, quantitative proteomic approach to identify the potential substrates of the deubiquitinase USP9X. In particular, TTK was identified as a potential substrate of USP9X. We demonstrated that USP9X directly interacted with and stabilized TTK. Knockdown of USP9X increased TTK ubiquitination, decreased TTK protein level, and suppressed cell proliferation, migration, invasion and tumor growth, demonstrating a phenotype similar to TTK depletion. Consistently, significant correlation between USP9X and TTK expression levels was observed in human NSCLC tissues, where they are over-expressed compared with matched adjacent normal tissues. Taken together, these results suggested that the USP9X-TTK axis could be considered as a potential therapeutic target for NSCLC.

Results

Identification of TTK as a candidate substrate for USP9X

To identify the candidate substrates of USP9X, we performed a proteome-wide screening experiment. Briefly, USP9X was knocked down in A549 cells with 3 shRNAs targeting different regions of the USP9X transcript. Expressions of endogenous proteins treated with shUSP9X were compared to the control shRNA-treated cells using a tandem mass tags (TMT)-based quantitative proteomic approach (Figure ). As a result, a total of 7471 proteins were identified in the proteomic experiment, and 22 proteins were remarkably down regulated in USP9X knockdown cells (fold change > 1.5, students' t test P value < 0.05; Figure and Table ), which could be the candidate substrates of deubiquitinase USP9X. Effective knockdown of USP9X was verified by Western blotting (Figure ). Moreover, the down-regulation of USP9X in the shUSP9X-treated group was further confirmed by the TMT-based proteomic quantification results. Among these significantly down-regulated proteins, TTK was selected as a potential substrate of USP9X (the ratio between the protein intensities of the shUSP9X and control shRNA samples was 0.62, P value = 0.02, Table ). In addition, the expression of TTK showed a significant correlation with that of USP9X, indicating that down-regulation of TTK was strongly correlated with decreased USP9X expression (Figure ). Several cell cycle-associated proteins, including CLASPIN, XIAP, SURVIVIN, and centrosome protein CEP131, have recently been reported as the substrates of USP9X 28, 30-32. These results indicated that TTK, as one of the spindle checkpoint proteins, can also be considered as a potential substrate of USP9X.

USP9X stabilizes the TTK protein by deubiquitination

To test the hypothesis that TTK is a direct substrate for USP9X, we first examined the interaction between USP9X and TTK. By constructing a flag tagged-TTK construct, a transfection/co-immunoprecipitation experiment was performed in 293T cells, which revealed that TTK interacts with USP9X (Figure ). We also detected the interaction of endogenous USP9X and endogenous TTK by performing immunoprecipitation experiments using USP9X antibody or IgG control from 293T cell protein extracts (Figure ). Furthermore, in order to validate the correlation between USP9X and TTK in NSCLC cells, we exploited siRNA-mediated depletion of USP9X in A549 cells using two different siRNAs targeting USP9X mRNAs. Both USP9X siRNAs markedly reduced USP9X protein levels after 48 h transfection, and also resulted in decreased TTK levels (Figure ). Moreover, the effect of USP9X depletion on TTK levels was not only found in A549 cells but also observed in other NSCLC cell lines, including NCI-H1975, NCI-H125 and NCI-H3122 (Figure ). And, examination of the protein expression levels of USP9X and TTK in multiple lung cancer cell lines showed a positive correlation (Figure ). The effect of USP9X reduction on the level of TTK was observed in HeLa cells as well (Figure ).

To test whether USP9X regulates TTK through deubiquitination, haemegglutinin (HA)-Ub, Flag-TTK, and siRNAs targeting USP9X or control siRNAs (siCTRL) were co-transfected into 293T cells followed by treatment with the proteasome inhibitor MG132. The results suggested that knockdown of USP9X increased the ubiquitination of TTK protein (Figure ). Furthermore, to test whether USP9X regulates TTK by deubiquitinating ubiquitin chains through K48 or K63 linkages, ubiquitin mutant vectors K48 and K63 (containing arginine substitutions of all of its lysine residues except those at positions 48 and 63, respectively), Flag-TTK, and siRNAs targeting USP9X or control siRNAs (siCTRL) were co-transfected into 293T cells followed by treatment with the proteasome inhibitor MG132. USP9X-mediated polyubiquitination of TTK was significantly increased in the presence of ubiquitin K48 plasmids, which indicated USP9X could efficiently remove the Lys 48 rather than Lys 63 type of the ubiquitin chain on TTK (Figure and Figure ). To consolidate the finding that USP9X protects TTK from ubiquitination-dependent proteasomal degradation, we directly used the proteasome inhibitor Bortezomib to treat cells transfected with either control siRNA or USP9X-targeting siRNA. Although endogenous USP9X was effectively decreased after siRNA transfection as indicated, TTK protein level was still increased in both of the transfected cells in a manner very similar to p27, which is a known substrate of the proteasomal degradation pathway (Figure ) 33.

To further explore whether USP9X enhances TTK stability, we transfected either control siRNA or USP9X-targeting siRNA into A549 cells to determine the half-lives of TTK. After 48 h transfection, the protein synthesis inhibitor cycloheximide (CHX) was added and cells were harvested at the indicated time points for immunoblotting. As expected, following the depletion of USP9X, TTK protein levels rapidly decreased (Figure ). Similar results were obtained in A549 cells stably expressing shGFP and shUSP9X (Figure ). We then went on to test whether stabilization of TTK by USP9X is achieved through the deubiquitination enzymatic activity of USP9X 30. Briefly, wild type USP9X or catalytically inactive C1566S mutant were transfected in HeLa cells and cycloheximide was added 36 h post-transfection. Overexpression of wild-type USP9X rather than the C1566S USP9X inactive mutant led to an increase in TTK protein level (Figure ). Similar results were obtained in 293FT cells (Figure ). The results outlined above suggested that USP9X interacts with and deubiquitinates TTK for its stabilization.

USP9X and TTK regulate cell migration, invasion, and colony formation

Overexpression of USP9X has been reported in NSCLC, resulting in tumorigenesis and poor prognosis 20, 34. To gain insight into the functional connection between USP9X and TTK on lung cancer migration and invasion, A549 cells were transfected with either control siRNA or two different siRNAs targeting USP9X, and subjected to wound-healing assays 48 h post-transfetion. Knockdown of USP9X/ TTK significantly delayed the cells from recovering the wounded area (Figure ). Western blotting experiments confirmed that TTK was significantly knocked down by the specific siRNAs (Figure ). We further performed invasion assays with A549 cells stably expressing shUSP9X and shTTK, respectively. Similarly, decreased expression of either USP9X or TTK remarkably reduced the number of invaded cells (Figure ). Therefore, the data above supported a functional consistency of USP9X and TTK in cell migration and invasion. We next performed colony formation assays to determine the impact of USP9X and TTK reduction. As shown in Figure and Figure , knockdown of USP9X or TTK dramatically decreased the colony numbers in A549 cells or Hela cells treated with shUSP9X or shTTK.

We next investigated the effect of WP1130, a partially selective inhibitor of deubiquitinating enzymes, directly inhibiting DUB activity of USP9X, USP5, USP14, and UCH37 35, in regulating TTK expression and cell proliferation. As expected, WP1130 decreased the colony forming ability of A549 cells in a dose-dependent manner (Figure ). Furthermore, we treated A549 cells with WP1130 and found that TTK expression levels were affected in a dose- and time-dependent manner (Figure ). To obtain better insights into the role of USP9X in TTK stability, HeLa cells were arrested in mitosis with nocodazole. Cycloheximide was then added to prevent further protein synthesis in the presence or absence of the USP9X inhibitor WP1130 and cells were harvested at the indicated time points. The results illuminated that WP1130 treatment destabilized the TTK proteins in mitosis cells (Figure ). Consistently, A549 cells became more sensitive to WP1130 with USP9X or TTK knockdown (Figure ).

USP9X and TTK modulate A549 cell proliferation and tumorigenesis

In order to uncover the function of USP9X/TTK in NSCLC carcinogenesis, we developed A549 cells stably expressing shUSP9X or shTTK (Figure ). Firstly, we examined whether USP9X or TTK could influence A549 cell proliferation. We seeded A549 cells with stable knockdown of USP9X or TTK, and measured the cell proliferation daily. Both USP9X-deficient and TTK-deficient A549 cells exhibited a much slower growth rate compared with shGFP A549 cells (Figure ). Then, to further figure out how USP9X/TTK functions in NSCLC tumorigenesis, A549 cells with stable knockdown of USP9X or TTK were injected into the right flank of nude mice, and the size of the resulted tumors was measured weekly. The measurements of the xenograft tumor size (Figure ) and weight (Figure ) demonstrated that the knockdown of USP9X or TTK suppressed tumor growth. The tumor proliferation curve showed that tumors originating from shUSP9X or shTTK-stable cells grew significantly slower than shGFP cells (Figure ). Finally, to test whether USP9X promoted A549 cell proliferation through TTK, we transiently overexpressed TTK in the shUSP9X cells. Our results demonstrated that transient overexpression of TTK could rescue cell proliferation suppression induced by USP9X knockdown (Figure ). Taken together, these results suggested that USP9X-mediated TTK stabilization plays an important role in tumorigenesis.

The expressions of USP9X and TTK are elevated and correlated in human NSCLC

Owing to its tumor promotion nature, we postulated that USP9X and TTK were both abnormally expressed in human cancers. To test this hypothesis, we examined the protein levels of USP9X and TTK in a tissue microarray by immunohistochemical staining (Figure ). High staining scores of USP9X and TTK were obtained in 39 out of the 55 (71%) NSCLC tumor samples. While only 11% (6 out of 55) for USP9X and 13% (7 out of 55) for TTK in the matched normal tissues showed high expression, respectively (Figure ). Statistical analysis revealed that there is a significantly positive correlation between USP9X/TTK expression and NSCLC. Meanwhile, about 56% of tumor samples with low USP9X (n =16) showed low TTK expression, while 82% of the high USP9X samples (n =39) exhibited high TTK expression (Figure ). Moreover, we found a trend between USP9X/TTK expression and pathologic differentiation. Although USP9X and TTK were not elevated in stage I tumors compared to normal tissues, they were elevated in tumors ranging from stage II to III (Figure ). These results suggested that the progressive NSCLC were highly associated with USP9X and TTK expressions.

Discussion

USP9X has been reported to be an oncogene or a tumor suppressor in different cancers primarily based on the context-specific function of its substrates 36. To shed light on the functional role of USP9X in lung cancer, we carried out TMT-based quantitative proteomics to identify substrate candidates of USP9X. Compared with a TMT quantitative siUSP9X proteome dataset in MCF-7 cells presented by Wang et al. 37, we identified 94% of the total proteins in their reported dataset, and there were 3265 more proteins identified in our study (Figure ). Consistent with the MCF-7 proteome, we identified the known USP9X substrate protein CEP55. Moreover, 96 additional down-regulated proteins (using criterion adapted from Wang et al.) were uniquely represented in our dataset. Therefore, the quantitative proteomic method in this study was highly sensitive and reproducible, thus allowing for identification of the low abundance substrate proteins of USP9X, such as kinases and transcriptional factors.

Among these down-regulated proteins, TTK was verified to have the same downregulation trend as USP9X (Figure ). Depletion of USP9X by siRNA led to reduced protein level and stability of TTK (Figure ). Immunoprecipitation assays showed that there was a direct interaction between USP9X and TTK (Figure ). Our experiments also demonstrated that a reduction in USP9X also led to a reduction of TTK stability in a proteasome-dependent manner (Figure ). Furthermore, we evidenced that depletion of USP9X only increased TTK K48 linkage poly-ubiquitination and had minimal effect on TTK K63 linkage poly-ubiquitination, further supporting that USP9X protects TTK from ubiquitination-mediated proteasome degradation (Figure and Figure ).

We noted that previous studies reported overexpression of USP9X in human NSCLC 20, 21. TTK mRNA level was also reported to be elevated 24, and TTK was considered as a tumor marker and therapeutic target for lung cancer (Y. N, Y. D, S. N, patent EP2102360A2). Therefore, we believed that the link between USP9X and TTK was one of multiple pathways that play important roles in cancer cells. We hypothesized that overexpression of USP9X elevates TTK levels, which in turn promotes progression of NSCLC. Although USP9X has been reported to have multiple substrates and participate in many cellular processes, our experimental results reveal that TTK is another substrate of USP9X and contributes to the function of USP9X. Suppression of USP9X or TTK expression caused significant proliferation, migration and growth inhibition of A549 cells (Figure and Figure ). These results suggested that USP9X may regulate cell proliferation, migration and growth through its control of TTK degradation. Transiently overexpressed TTK on shUSP9X cells could rescue the USP9X knockdown-induced cell proliferation suppression (Figure ), demonstrating that TTK could contribute to the effect of USP9X depletion.

The partially selective DUB inhibitor WP1130 triggers aggresome formation and directly inhibits DUB activity of USP9X, USP5, USP14 and UCH37 to increase tumor cell apoptosis 35. WP1130 treatment suppressed USP9X expression, resulting in transcription factor E-twenty-sis related gene (ERG) degradation and growth inhibition of ERG-positive prostate tumors 38. Previous studies stated that WP1130 increases cisplatin sensitivity in estrogen receptor-negative breast cancer cells through inhibition of USP9X 39. WP1130 becomes an effective tool in investigation of DUBs function in cancer. In our study, we confirmed that WP1130 reduced the effect of USP9X to stabilize TTK by cycloheximide-chase assays (Figure ). We additionally found that USP9X inhibitor WP1130 decreased cell proliferation of A549 cells (Figure ). Meanwhile, we observed that A549 cells depleted of either USP9X or TTK showed increased sensitivity towards WP1130 treatments (Figure ), which indicates the impact of TTK on the function of USP9X. To further investigate the role of USP9X/TTK in tumorigenesis of NSCLC, we transplanted USP9X- or TTK-depleted A549 cells into the right flank region of nude mice, and observed greatly suppressed tumor growth in nude mice with A549 cells expressing lower levels of USP9X or TTK (Figure ).

In addition, we examined the protein levels of USP9X and TTK in clinical samples of NSCLC. Our results revealed that both USP9X and TTK expressions were elevated in NSCLC tissues and positively correlated to NSCLC progression. In conclusion, this study identified USP9X as a deubiquitinase of TTK, and USP9X promotes tumorigenesis of NSCLC through stabilizing TTK (Figure ).

Future studies are warranted to investigate the physiological functions of USP9X and its synergy with other DUBs to control TTK in vivo. Sufficient elucidation of TTK function in cell proliferation and tumor growth is also valuable for understanding the post-translational regulatory spectrum in cancer cells.

Materials and methods

Plasmids, siRNAs and antibodies

Plasmids expressing the full length USP9X (pEF-DEST51 V5-tagged mouse USP9X) or the catalytically inactive mutant (C1566S) were kindly provided by professor Stephen Wood, Eskitis Institute for Cell and Molecular Therapies, Griffith University, Brisbane, Queensland, Australia. Human TTK was cloned into pCMV-Flag. All siRNA sequences were synthesized by GenePharma (Shanghai, China). Two out of three siRNAs of USP9X were previously described 30: 5'- AGAAAUCGCUGGUAUAAAUUU -3' (siUSP9X_1), 5'- GCAGUGAGUGGCUGGAAGUTT -3' (siUSP9X_2). The other one is 5'- GGACUUCUUUGAAAGUAAUTT -3' (siUSP9X_3). The siRNA target sequences of TTK are: 5'- GGAUUUAAGUGGCAGAGAATT -3' (siTTK_1), 5'- GGUCGUUACAGUCAAGCAATT - 3' (siTTK_2) and 5'- CCAGAGGACAGACUACUAATT -3' (siTTK_3). Plasmids and siRNAs were transfected with Lipofectamine 3000 (Thermo, L3000015). All siRNAs were used at 60 nM. The following antibodies were used: β-ACTIN (Cell Signaling, 3700), USP9X (Cell Signaling, 14898), TTK (Millipore, 05-682), p27 (Cell Signaling, 3686), HA (Sigma-Aldrich, H6908), Flag (Sigma-Aldrich, SAB4301135), His (Sigma-Aldrich, H1029), IgG (Cell Signaling, 2729).

Cell culture and drug treatments

A549, 293T, HeLa, NCI-H3122, NCI-H125, NCI-H1975 and 293FT cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China), and A549, 293T, HeLa and 293FT were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco, 26400044) and 1% penicillin/streptomycin. NCI-H3122, NCI-H125 and NCI-H1975 were maintained in Roswell Park Memorial Institute 1640 (RPMI 1640). MG132 (Sigma-Aldrich, C2211) was used at 10 μM, cycloheximide (CHX; Targetmol, T1225) was used at 100 μg/mL unless otherwise specified, nocodazole (Noc; Sigma-Aldrich, M1404) at 100 ng/mL, WP1130 (SelleckChem, S2243) at 5 μM unless otherwise specified, and Bortezomib (SelleckChem, S1013) at 10 μM. For synchronization in mitosis, A549 and HeLa cells were treated with nocodazole for 16 h.

shRNA constructs and Lentiviral production

The target sequences of USP9X shRNAs: 5'- CGACCCTAAACGTAGACATTA - 3' (shUSP9X_1), 5'-CGATTCTTCAAAGCTGTGAAT - 3' (shUSP9X_2) and 5' -CCACCTCAAACCAAGGATCAA -3' (shUSP9X_3). The target sequences of TTK shRNAs: 5' -GCACAATTTGAACTGTCACAA -3' (shTTK_1) and 5' -GCCAACTTGTTGGTCTGAATT - 3' (shTTK_2). The shRNAs of USP9X and TTK were constructed in pLKO.1 vector. Plasmid shRNAs in the PLKO.1 vector were co-transfected into 293T cells along with two assistant vectors containing the vsvg and Δ8.9 genes. Lentivirus was harvested at 24 h and 48 h after transfection and subsequently clarified by a 0.45 μm filter. 10 μg/mL polybrene was added to the viruses before infecting cells, and stable cell lines were selected with treatment of 2 μg/mL puromycin for one week.

Cell viability and proliferation analysis

Cell viability was analyzed using Cell Counting Kit-8 (CCK-8; Beyotime, C0038). Cells were seeded into 96-well plates (7000 cells/well). After treatment with different concentrations of WP1130, the optical density (OD) at 490 nm for each well was obtained using a microplate reader (BIO-RAD Model 3550). Cell proliferation was also analyzed with CCK-8 assay. Cells were seeded into 96-well plates (500 cells/well), then incubated at 37 °C for the next 6 days. From day 1 to day 7, CCK-8 assay was performed conforming to the manufacturer's instructions. The OD value at 450 nm for each well was measured.

Immunoprecipitation and immunoblotting

Immunoprecipitation and immunoblotting were performed as described previously 40.

Cohort and immunohistochemistry

The immunohistochemistry was done by Shanghai Outdo Biotech Company, Shanghai, China. 55 pairs of NSCLC tissues with matched adjacent normal tissues were analyzed. Incubation with antibodies against USP9X (1:100 dilution, Cell Signaling, #14898) or TTK (1:1000 dilution, Merck Millipore, MABE571) was carried out at 4 ºC overnight. Photomicrographs were taken with the ScanScope XT (Leica, CA, USA). Slides were scored by a certified pathologist blinded to the clinicopathological data of the samples. Each sample was assigned a score according to the intensity of the staining (no staining = 0, weak staining = 0-1, moderate staining = 1-2, strong staining = 2-3) and the percentage of positive cells (0% = 0, 1-24% = 1, 25-49% = 2, 50-74% = 3, 75-100% = 4) as previously reported 41. The final score was obtained by multiplying the intensity score with the percentage of positive cells. Samples exhibiting scores < 6 were called USP9X/TTK low, and samples showing scores 6-12 were summarized as USP9X/TTK high. The correlation between TTK and USP9X, and the correlation of TTK or USP9X with tissue type were verified by the chi-square test.

Colony formation assay

A549 cells stably expressing shRNAs targeting USP9X or TTK were plated onto 6-well plates (500 cells/well). After the cells had incubated for 14 days and their colonies were visible to the naked eye, each well was washed three times with PBS, fixed with methanol for 10 min, and then stained with 1% crystal violet solution for 15 min. The numbers of colonies per well were manually counted. The experiments were independently triplicated. For the WP1130 treatment assay, the indicated concentration of WP1130 was added to the cells after plating (500 cells/well). After 14 days, the cells were washed with PBS, fixed with methanol, stained with 1% crystal violet and counted.

Cell migration and invasion assay

A549 cells treated with different siRNAs targeting USP9X or TTK were plated in 96-well plates for wound-healing assays. The plates were photographed by IncuCyte ZooM (Essen BioScience, Michigan, USA). For the transwell assays (Costar, #3422), A549 cells stably expressing shRNAs were placed on the upper layer of a cell-permeable membrane and media containing 10% FBS was placed in the lower chamber. After incubation for 20 h, the cells that had migrated through the membrane were stained and counted.

Tumor xenografts

The animal experiment was carried out by the animal experimental center in Shanghai University of Traditional Chinese Medicine. A549 cells stably expressing shRNAs (5×106) were suspended in 100 μL PBS and subcutaneously injected into the right flank region of nude mice (4-6 weeks old). Tumor sizes were measured every 7 days with a caliper and tumor volumes were determined with the formula: L × W2 × 0.5, where L is the longest diameter and W is the shortest diameter. 4 weeks later, all mice were sacrificed and the tumors were harvested, photographed and weighed. All treatments were administered according to the guidelines of the Institution Animal Care and Use Committee and all protocols were approved by the Ethics Committee of Shanghai University of Traditional Chinese Medicine.

Protein extraction and filter-aided sample preparation

The A549 cells were washed three times with cold PBS and the cells were lysed in SDT lysis buffer (0.2% SDS (m/v), 100 mM DTT, 100 mM Tris, pH=7.6). The lysates were homogenized with sonication and centrifuged at 15,000 ×g for 30 min. The supernatants were collected, and protein concentration was determined by tryptophan fluorescence emission assay as described previously 42. FASP was performed as described previously with slight modifications 43, using 50 mM triethylammonium bicarbonate (TEAB) instead of ABC buffer for compatibility with TMT labeling.

TMT labeling

Peptides were reconstituted in 100 mM TEAB and their concentration determined by the BCA assay. 3 shGFPs control samples and 3 USP9X shRNAs knockdown samples (50 μg, each) were labelled by TMT 6-plex reagents (0.8 mg, each) (Thermo Fisher Scientific) with reporters at m/z = 126, 127, 128, 129, 130, 131, respectively. Each sample containing 50 μg of peptide in 50 μL volume of TEAB buffer was combined with 41 μL of its respective 6-plex TMT reagent and incubated for 1 h at room temperature. Then, 8 μL of 5% hydroxylamine was added to the sample and incubated for 15 min to quench the reaction. Equal amounts of each TMT-labelled sample were combined in new microcentrifuge tubes and dried down using a SpeedVac.

High pH fractionation

Half of the TMT-labeled peptides mixture was fractionated using a Waters XBridge BEH130 C18 3.5 μm 2.1 × 150 mm column on an Agilent 1290 HPLC operating at 0.2 mL/min. Buffer A consisted of 10 mM ammonium formate and buffer B consisted of 10 mM ammonium formate with 90% acetonitrile; both buffers were adjusted to pH 10 with ammonium hydroxide as described previously 44. A CBS-B programed multifunction automatic fraction-collecting instrument (Huxi instrument, Shanghai, China) was coupled to the HPLC and used to collect eluted peptides. The total number of fractions collected was 36 and concatenated to 19 (pooling equal interval RPLC fractions). Ammonium hydroxide and/or ammonium formate were evaporated in a SpeedVac.

Nanoflow liquid chromatography tandem mass spectrometry

All experiments were performed on an Orbitrap Fusion Lumos mass spectrometer with an ancillary nanoLC (EASY-nLC II; Thermo Fisher Scientific). Peptides were loaded on a self-packed column (75 μm × 150 mm, 3 μm ReproSil-Pur C18 beads, 120 Å, Dr. Maisch GmbH, Ammerbuch, Germany) and separated with a 120 min gradient for fractionation of samples and 180 min gradient for total samples at a flow rate of 300 nL/min. Solvent A was 100% H2O, and 0.1% formic acid; solvent B was 100% acetonitrile, and 0.1% formic acid. The Orbitrap Fusion Lumos was programmed in the data-dependent acquisition mode. An MS1 survey scan of 350-1500 m/z in the Orbitrap at a resolution of 60,000 was collected with an AGC target of 400,000 and maximum injection time of 20 ms. Precursor ions were filtered according to monoisotopic precursor selection, charge state (+ 2 to + 7), and dynamic exclusion (40 s with a ±10 ppm window). Then, the most intense precursors were subjected to HCD fragmentation with a duty cycle of 3 s. The instrument parameters were set as follows: 32% normalized collision energy, 30,000 resolution, 50,000 AGC target, 50 ms maximum injection time, 100 Da first mass, inject ions for all available parallelizable time enabled, 1.6 m/z isolation width.

MS raw data

The mass spectrometry proteomics data have been deposited to the ProteomeXchange 45 Consortium via the PRIDE 46 partner repository with the dataset identifier PXD007726.