Advanced cutaneous melanoma is one of the most challenging cancers to treat because of its high plasticity, metastatic potential, and resistance to treatment. New targeted therapies and immunotherapies have shown remarkable clinical efficacy. However, such treatments are limited to a subset of patients and relapses often occur, warranting validation of novel targeted therapies. Posttranslational modification of proteins by ubiquitin coordinates essential cellular functions, including ubiquitin-proteasome system (UPS) function and protein homeostasis. Deubiquitinating enzymes (DUB) have been associated to multiple diseases, including cancer. However, their exact involvement in melanoma development and therapeutic resistance remains poorly understood. Using a DUB trap assay to label cellular active DUBs, we have observed an increased activity of the proteasome-associated DUB, USP14 (Ubiquitin-specific peptidase 14) in melanoma cells compared with melanocytes. Our survey of public gene expression databases indicates that high expression of USP14 correlates with melanoma progression and with a poorer survival rate in metastatic melanoma patients. Knockdown or pharmacologic inhibition of USP14 dramatically impairs viability of melanoma cells irrespective of the mutational status of BRAF, NRAS, or TP53 and their transcriptional cell state, and overcomes resistance to MAPK-targeting therapies both in vitro and in human melanoma xenografted mice. At the molecular level, we find that inhibition of USP14 rapidly triggers accumulation of poly-ubiquitinated proteins and chaperones, mitochondrial dysfunction, ER stress, and a ROS production leading to a caspase-independent cell death. Our results provide a rationale for targeting the proteasome-associated DUB USP14 to treat and combat melanomas. Mol Cancer Ther; 17(7); 1416–29. ©2018 AACR.
Cutaneous melanoma is an aggressive type of skin cancer whose incidence increases worldwide. It is recognized for its propensity for early and extensive metastatic spread and seen as one of the most virulent and refractory form of human advanced cancers. Melanoma develops from the pigment-producing melanocytes in the epidermis (1). Tumor progression is accompanied by driver mutations affecting the BRAF and NRAS genes (in 50% and 20% of melanoma, respectively) leading to constitutive activation of the MEK/ERK MAPK pathway. Other genetic and epigenetic changes, as well as the tumor microenvironment, affect the survival and proliferation of melanoma cells by altering several signaling pathways. The majority of melanomas undergo an epithelial-to-mesenchymal (EMT)-like transition to acquire a mesenchymal motile phenotype associated with a drug-tolerant state and metastatic abilities (1–3). Melanomas also exhibit a high level of differentiation state plasticity and intratumor heterogeneity, which is an obstacle to therapeutic success and a source of relapse (4–6). Immunotherapies and therapies targeting mutant BRAF and MEK have led to improved survival in patients with metastatic disease (2, 7). Nevertheless, drug resistance to BRAF/MEK inhibition invariably develops within 6 to 12 months of treatment. Relapses are generally associated with acquired resistance linked to NRAS mutations that reactivate the MAPK pathway and activation of parallel signaling pathways such as the PI3K/AKT pathway (2, 5). The long-term prognosis of metastatic melanoma is still very poor for most patients, underscoring the need for novel therapeutic options (2, 8). Beside, no targeted treatments currently exist for patients whose tumors are wild-type for BRAF, including NRAS-mutant melanomas.
In cancer cells, accumulation of genetic mutations, aneuploidy, and chromosomal alterations, together with a high proliferative index, increases protein synthesis and cell addiction to mechanisms controlling protein homeostasis, including the heat shock and unfolded protein responses, autophagy, and endo-lysosomal degradative pathways, and the ubiquitin-proteasome system (UPS; refs. 9, 10). Posttranslational modification of proteins by ubiquitin represents a complex signaling system that coordinates essential cellular functions and protein homeostasis (11). While mono-ubiquitination is rather involved in the modification of protein activity and intracellular transport, poly-ubiquitination, with a K48 linkage, is the main mechanism responsible for the degradation of cellular proteins by the UPS system. This system encompasses a proteolytic machinery (the 26S proteasome) and a cascade of enzymes (E1, E2, and E3) capable of activating ubiquitin residues and adding them to the target proteins (12). Illustrating the importance of the UPS in melanoma, high levels of Skp2, an F-box protein of the ubiquitin ligase complex SCF, have been associated with p53-dependent proliferation (13), and inactivation of the E3 ligase FBXW7 leads to sustained accumulation of NOTCH1, thereby accelerating tumor development (14). We have also shown that the matricellular protein SPARC controls a signaling circuit from the tumor microenvironment leading to proteasome-mediated degradation of p53 via the E3 ligase MDM2, which confers survival advantage to melanoma cells (15). The enzymatic reaction, which opposes the conjugation of ubiquitin by E3 ligases, consists of the deubiquitination by deubiquitinating enzymes (DeUBiquitinases or DUBs). DUBs represent ubiquitin-specific cysteine proteases that can cleave one or more ubiquitin molecules on target proteins, or even the entire poly-ubiquitin chain (16, 17). The expression or abnormal activity of DUBs has been demonstrated in pathologic situations, including cancer and their druggable activity, is now considered for development of novel molecular therapies (18, 19). Several DUBs have been linked to melanoma progression and response to therapy, including USP13 (20), USP5 (21), and recently, USP9X as an important regulator of NRAS expression (22). Some DUBs, such as UCHL5 and USP14 (Ubiquitin-specific peptidase 14) are direct components of the proteasome, and their pharmacologic targeting by compounds such as b-AP15 and VLX1570 has a major impact on cancer proteostasis and immunoreactivity (23–25). USP14 is an essential regulator of the proteasome (26, 27) that control the deubiquitination of poly-ubiquitinated substrates addressed to the proteasome (28). The function of USP14 in promoting cell growth and viability in diverse type of cancers, such as acute myeloid leukemia (23), multiple myeloma (29), squamous and colon carcinoma (23), and breast cancer (30), has been recognized. In this context, the recent observation that Akt activates USP14 to control the ubiquitin–proteasome system (31) further underlies the involvement of USP14 in tumorigenesis. However, its role in the malignancy of melanoma remains poorly understood.
In this study, using a combination of biochemical, genetic, and pharmacologic approaches, we found that targeting USP14 impairs melanoma cell viability irrespective of the mutational status, and overcomes resistance to targeted therapies in vitro and in vivo. Mechanistically, inhibition of USP14 resulted in a caspase-independent cell death associated with accumulation of poly-ubiquitinated proteins and chaperones, ROS production, mitochondrial dysfunction, and ER stress. Our results show that the proteasome-associated DUB USP14 represents a promising therapeutic target in melanoma.
Materials and Methods
Cell and reagents
The human melanoma cell lines A375, 1205lu, and MeWo were from ATCC. HMVII cells were from European Collection of Authenticated Cell Cultures (ECACC). WM164 and WM266–4 cells were purchased from Rockland. SBCL2, WM793, 451lu, Mel1617 cells and their BRAF inhibitor–resistant derivatives (451luR and Mel1617R) were provided by Drs. M. Herlyn and J. Villanueva (The Wistar Institute, Philadelphia, USA) and have been described previously (33). M229, M238, and M249 cells and their BRAF inhibitor–resistant sublines were from Dr. R. Lo (UCLA Dermatology, Los Angeles, USA) and have been characterized previously (34). 501Mel cells were a gift from Dr. R. Halaban (Yale University School of Medicine, New Haven, USA) and were described elsewhere (3). All melanoma cell lines were used within 6 months between resuscitation and experimentation. To guarantee cell line authenticity, cell lines were expanded and frozen at low passage after their receipt from original stocks, used for a limited number of passages after thawing, and routinely tested for the expression of melanocyte-lineage proteins such as Microphthalmia-Associated Transcription Factor (MITF). All cell lines were routinely tested for the absence of mycoplasma by PCR. A375DR double resistant cell line variants were generated from A375 cells by repeated exposure to BRAF inhibitor PLX4032 (Selleckchem) and ERK inhibitor SCH772984 (Selleckchem) until the onset of resistance. Patient melanoma cells (Pt#1 and Pt#2) were a kind gift from Dr. R. Ballotti (INSERM U1065, Nice, France) and have been described previously (3). Melanoma cells were cultured in DMEM supplemented with 7% FBS (HyClone). Human primary epidermal melanocytes were isolated from foreskin and maintained as described previously (32). For in vivo bioluminescence imaging, 451luR-Luc+ cells were obtained by lentiviral transduction (pLenti6/V5-luciferase; Thermo Fischer Scientific) and blasticidin selection (2 μg/mL). All cell cultures were grown at 37°C under 5% CO2.
Primers and culture reagents were purchased from Thermo Fischer Scientific. b-AP15 and VLX1570 were from Merck Millipore and MedChem Express, respectively. Bortezomib was from Selleckchem. QVD-OPH was from ApexBio Technology. Staurosporin, MG-132, N-Acetyl-L-cysteine (NAC), H2O2, and all other reagents were purchased from Sigma-Aldrich unless otherwise stated.
Antibodies used were USP14, HSP60, HSP70, p53, p21, ERK2 (Santa Cruz Biotechnology), K48-linkage–specific polyubiquitin (D9D5), JNK, phospho-JNK (Thr183/Tyr185), p38, phospho-p38 (Thr180/Tyr182), phospho-Rb (Ser807/Ser811), Caspase 8, CHOP, Tubulin α (Cell Signaling Technology), UCHL5, USP5, USP7 (Bethyl Laboratories), K63-linkage specific Polyubiquitin (Abcam), and Bip/GRP78 (BD Biosciences). Peroxidase-conjugated anti-rabbit antibodies were from Cell Signaling Technology. Peroxidase-conjugated anti-mouse and anti-goat antibodies were from Dako.
RNAi studies, transfection, and infection studies
siRNAs were purchased from Dharmacon (Thermo Fisher Scientific). Transfection of siRNA was carried out using Lipofectamine RNAiMAX (Thermo Fisher Scientific) at a final concentration of 25 or 50 nmol/L. Cells were assayed 3 or 6 days posttransfection. USP14-Flag adenovirus was purchased from Vigene Biosciences. Cells were infected for 2 days in DMEM supplemented with 7% FBS prior treatment.
Cell lysis and immunoblot analysis
Melanoma cells were harvested as described before (35). Cells were lysed at 4°C in RIPA buffer (Millipore) supplemented with Pierce Protease and Phosphatase Inhibitor Mini Tablets, and briefly sonicated. Cell lysates were cleared at 16,000 × g for 15 minutes at 4°C. Whole-cell lysates were subjected to SDS-PAGE and immunoblot analysis as described previously (35).
DUB trap assay
Cells were lysed in ice-cold buffer containing 50 mmol/L Tris (pH 7.4), 5 mmol/L MgCl2, 250 mmol/L sucrose, 1 mmol/L DTT, 2 mmol/L ATP, and 1 mmol/L PMSF and mild sonication. Lysates were cleared by centrifugation and 30 μg of protein extracts were incubated for 15 minutes at 37°C with 0.5 μmol/L HA-Ub-VS (Boston Biochem). After boiling in reducing sample buffer, labeled cell lysates were subjected to immunoblot analysis.
Proteasome activity assay
Proteasome activity was determined as described previously (36). Cells were lysed for 30 minutes at 4°C in ATP-containing lysis buffer (50 mmol/L HEPES pH 7.8, 5 mmol/L ATP, 0.5 mmol/L DTT, 5 mmol/L MgCl2, 0.2% Triton X-100). Cell lysates were cleared at 16,000 × g for 15 minutes at 4°C. Ten micrograms of proteins from each sample were incubated in 96-well plates with 0.1 mmol/L of Z-Leu-Leu-Glu-AMC, Suc-Leu-Leu-Val-Tyr-AMC, and Ac-Arg-Leu-Arg-AMC fluorogenic substrates (Enzo Life Sciences) to measure caspase-like, chymotrypsin-like, and trypsin activities, respectively. Fluorescence intensity was measured during 2 hours by following emission at 460 nm (excitation at 390 nm).
Aggresome staining was performed using the Proteostat Aggresome detection kit (Enzo Life Sciences) according to the manufacturer's instructions. Analysis was performed by flow cytometry.
Cell proliferation was measured by a MTS conversion assay using the CellTiter 96 Aqueous Non-Radioactive Cell Proliferation kit (Promega) according to the manufacturer's instructions (5 × 103 cells/well). Alternatively, cell growth was assessed by crystal violet staining on cells seeded in 24-well plates for the indicated time. After treatment, cells were fixed in PFA 3% during 20 minutes, washed with PBS three times, and stained with crystal violet 0.4% in ethanol 20% for 30 minutes. For real-time analysis of cell growth using an automated IncuCyte Zoom imaging system (Essen Bioscience), cells were plated in triplicate in complete medium on 96-well plates (5 × 103 cells/well). Phase contrast images were taken every hour over a 3-day period. Growth curves were generated using the IncuCyte cell proliferation assay software based on cell confluence.
Analysis of apoptosis and cell cycle by flow cytometry
Cell-cycle analysis was performed by flow cytometry analysis of propidium iodide (PI)-stained cells. Following permeabilization in ice-cold ethanol 70%, cells were stained with 40 μg/mL PI in PBS supplemented with 100 μg/mL RNAse A before analysis using a BD FACSCanto cytometer. Cell death was evaluated by flow cytometry following staining with Annexin-V-FITC and PI (eBioscience) as described previously (37). Alternatively, cells were seeded in 96-well plates (5 × 103 cells/well) overnight at 37°C. A mixture of compounds and Cytotox Green Reagent (Essen Bioscience) in complete culture medium was then added to the cells. Cell death was monitored in real time for 72 hours with an IncuCyte imaging system. Experiments were carried out in triplicate.
Measurement of ROS production
To measure ROS levels, cells were stained with 10 μmol/L of the redox-sensitive dye CM-H2DCFDA (Thermo Fischer Scientific) in PBS for 30 minutes at 37°C. Cells were washed with PBS and resuspended in PBS 5 mmol/L EDTA/1% BSA. ROS production was analyzed using a MACSQuant Analyzer 10 cytometer (Miltenyi Biotec).
Measurement of mitochondrial membrane potential
Following treatment, cells were stained with 1 μmol/L tetramethylrhodamine (TMRE) in media for 15 minutes at 37°C. Cells were washed with PBS and resuspended in PBS, 5 mmol/L EDTA/1% BSA. MMP levels were analyzed using a MACSQuant Analyzer 10 cytometer (Miltenyi Biotec).
Tumors frozen-section in optimal cutting temperature were sliced using a Cryostat CM350 (8 μm) and stained using the Apoptag Red In Situ kit (Millipore) according to the manufacturer's instructions. Images were captured on a Leica widefield microscope (Leica Microsystems).
Real-time quantitative PCR
Total RNA was extracted from cell samples using Nucleospin RNAII kit (Macherey-Nagel). Reverse transcription was performed on 1 μg of total RNA in a volume of 20 μL using High capacity cDNA Reverse Transcription kit (Applied Biosystems). Quantitative PCR was performed on 20 ng cDNA samples, in sealed 96-well microtiter plates using the Platinum SYBR Green qPCR Supermix-UDG w/ROX (Life Technologies) with the StepOnePlus System (Applied Biosystems). Relative mRNA levels were determined using the 2ΔΔCt method and ACTB and PPIA as housekeeping genes. Values are the mean of duplicates and are representative of two independent experiments.
Gene expression analysis from human databases
Publicly available gene expression data sets from Gene Expression Omnibus (GEO) database were used to analyze USP14 levels in melanoma progression (GSE3189) and patient outcome (GSE19234). Normalized data were analyzed using GraphPad Prism V5.0b software (GraphPad).
In vivo experiments
All mouse experiments were carried out in accordance with the Institutional Animal Care and the local ethics committee. For human melanoma xenografts, 5-week-old female athymic (nu/nu) mice (Harlan) were subcutaneously injected with 1 × 106 BRAF inhibitor–resistant 451luR melanoma cells engineered to express a luciferase reporter (451luR-Luc+ cells) in 100 μL of PBS. After 3 days, mice were injected every 3 days intraperitoneally with vehicle or 10 mg/kg b-AP15 in 90/1/9 mix of Labrafil/Tween 80/DMA. At the indicated times, mice were anesthetized and injected intraperitoneally with 50 mg/kg d-Luciferin (Perkin Elmer) in PBS. Images were acquired using a Photon Imager (Biospace Lab) system and data analyzed with the M3Vision software (Biospace Lab). Tumor growth was monitored and quantified using bioluminescence imaging (BLI). The total numbers of photons per second per steradian per square centimeter were recorded. For BLI plots, photon flux was calculated for each mouse by using a rectangular region of interest encompassing the thorax of the mouse in a prone position. This value was normalized to the value obtained immediately after injection (15 minutes), so that all mice had an arbitrary starting BLI signal of 100.
Unless otherwise stated, all experiments were repeated at least three times and representative data/images are shown. Statistical analysis was performed using the GraphPad Prism software. All data are presented as mean ± SEM. For comparisons between two groups, P values were calculated using unpaired one-sided t test or Mann–Whitney test. Statistical significance of the in vivo experiment was calculated with the two-way ANOVA test. P values of 0.05 (*), 0.01 (**), and 0.001 (***) were considered statistically significant.
Expression and activity of USP14 in melanoma cells
To identify DUBs that may be involved in melanoma biology, we used a DUB trap assay with the probe HA-Ub-VS, which covalently labels with a HA-tagged ubiquitin molecule active DUBs in cell lysates (ref. 38; Fig. 1A). The DUB trap assays performed on melanocyte and melanoma cell lysates revealed an increased activity of several DUBs in melanoma cells compared with melanocytes. Anti-USP14 immunoblot analysis of the above DUB trap assays showed that compared with normal melanocytes, USP14 activity is significantly increased in melanoma cell lines 501Mel, 451lu, MeWo, and A375 (Fig. 1B). To further examine the role of USP14 in melanoma, we used the previously described USP14 inhibitor b-AP15 (23). b-AP15 efficiently blocked the activity of USP14 in a dose-dependent manner in A375 cells (Fig. 1C; Supplementary Fig. S1A), and in 1205lu, 501Mel, SKMel28, and 451lu cells (Fig. 1D). Notably, the DUB activity of UCHL5, another target of b-AP15 (29), was not affected in melanoma cells. Note that the activity of other DUBs, such as USP5 and USP7, were also not significantly affected by b-AP15 treatment (Supplementary Fig. S1A). Consistent with the effect of b-AP15, VLX1570, a selective inhibitor of USP14 (39, 40) also blocked USP14, but not UCHL5, activity in melanoma (Supplementary Fig. S1B and S1C).
To confirm USP14 as a potential target in melanoma, we assessed USP14 gene expression in annotated NCBI Gene Expression Omnibus (GEO) datasets comparing patients' tumors with their normal tissue counterparts and benign tumors. While USP14 levels were not significantly increased between normal skin samples and nevi, its expression was statistically increased in melanomas when compared with both normal skin and nevi (Fig. 1E). Interestingly, further analysis of public databases associates high USP14 expression with a lower probability of survival in patients with metastatic melanoma (Fig. 1F).
USP14 inhibition has a potent and broad anti-melanoma effect
We next examined the impact of USP14 inhibition on a collection of melanoma cell lines with diverse mutational status and metastatic potential (3, 4). These cell lines are also representative of the three major subtypes of melanoma based on the frequency of BRAF and NRAS mutations: mutant BRAF, mutant RAS, and wild-type. The exposure of 1205lu, 501Mel, and A375 cells to b-AP15 during 48-hour affected cell morphology (Supplementary Fig. S2A) and impaired cell proliferation in a dose-dependent manner (Fig. 2A). Using the same approach, we calculated the IC50 of b-AP15 for 22 melanoma cell lines and short-term cultures to be in a micromolar range (0.3–3.4 μmol/L; Table 1). Our data show that pharmacologic inhibition of USP14 by b-AP15 similarly reduced proliferation of mutant BRAF, mutant RAS, and wild-type melanoma cells. Treatment with b-AP15 also blocks the proliferation of melanoma cells irrespective of the mutational status of TP53, PTEN, or CDKN2A and of their invasive or proliferative transcriptional cell state. Real-time imaging further shows that the effect of b-AP15 on cell proliferation was rapid and compared with what was observed with the proteasome inhibitor bortezomib or with VLX1570 (Fig. 2B; Supplementary Fig. S2B). We next evaluated how b-AP15 affects melanoma clonogenic growth. As shown in Fig. 2C, 501Mel and A375 cells treated with b-AP15 were no longer capable of forming colonies when isolated, compared with untreated cells. Cell-cycle analysis performed on melanoma cells treated or not with b-AP15 for 24 hours revealed that USP14 inhibition altered cell-cycle progression and increased cell death as indicated by the appearance of a sub-G1 cell population with reduced DNA content (9% and 5% for 1205lu and A375 cells, respectively; Fig. 2D). For comparison, BRAF inhibitor vemurafenib (PLX4032) caused cell-cycle arrest in the G0–G1 phase.
|.||Type .||Mutation(s) .||Resistance .||IC50 (μmol/L) .|
|.||Type .||Mutation(s) .||Resistance .||IC50 (μmol/L) .|
NOTE: IC50 (μmol/L) of b-AP15 treatment on melanoma cell proliferation was determined after 48 hours by a MTS conversion assay.
Abbreviations: MET, metastasis; RGP, radial growth phase; RTK, receptor tyrosine kinase; VGP, vertical growth phase.
aGene mutation or alteration.
The impact of b-AP15 on melanoma cell death was confirmed through real-time imaging of cell toxicity (Fig. 2E) and with a flow cytometric analysis of Annexin-V-FITC/PI labeling of A375 cells treated or not with b-AP15 (Fig. 2F). Data showed a significant reduction of the viability of A375 cells exposed to b-AP15 (45%) compared with solvent effect (93%), with increase in both early and late apoptotic populations. As a control, treatment with the cell death inducer staurosporine led to a similar decrease in cell viability (37%). Importantly, melanoma cell death was induced to similar levels following incubation with VLX1570 (Supplementary Fig. S2C and S2D). At the molecular level, immunoblot analysis showed that USP14 targeting in 1205lu cells altered the expression or phosphorylation of proteins related to cell proliferation and apoptosis. Treatment with b-AP15 increased levels of the cell-cycle inhibitor p21CIP1, while reducing the levels of phosphorylated Rb proteins. After 24 hours of treatment, the drug caused the appearance of active fragments of caspase 3 and the cleavage of its nuclear substrate PARP (Fig. 2G). Interestingly, washout experiments performed on A375 and 501Mel cells further showed that a short exposure to b-AP15 (1 hour) followed by 23-hour incubation in drug-free medium had an anti-melanoma action similar to a full exposure to b-AP15 for 24 hours (Supplementary Fig. S2E and S2F), indicating that the cytotoxic action of b-AP15 in melanoma is rapid and irreversible, consistent with a previous study (41). Finally, we found that adenoviral-mediated overexpression of USP14 significantly antagonized the cleavage of PARP induced by an optimal dose of b-AP15 (2 μmol/L), suggesting that the cytotoxicity of b-AP15 on melanoma cells was largely not due to an off-target effect (Supplementary Fig. S3). Together, these observations demonstrate that the pharmacologic inhibition of USP14 has a potent anti-melanoma effect irrespective of mutational status of oncogenes and gene expression signatures.
Depletion of USP14 in melanoma reduces cell survival
We next used a genetic approach based on siRNAs targeting USP14 to study the effects of USP14 depletion on melanoma cell survival. Compared with nontargeting siRNA, two siRNA sequences (siUSP14 #1 and # 2) targeting USP14 efficiently decreased USP14 expression and similarly decreased the phosphorylation of Rb following 3 days of transfection in 1205lu cells (Fig. 3A). Real-time imaging shows that USP14 knockdown impaired the proliferation of melanoma cell lines A375 and 501Mel (Fig. 3B; Supplementary Fig. S4A). We then studied the impact of USP14 depletion on melanoma focus formation. Compared with cells transfected with control siRNA, A375 and 501Mel cells transfected with USP14 siRNA were no longer capable of forming colonies after 7 days (Fig. 3C; Supplementary Fig. S4B and S4C). Consistent with this, cell-cycle analysis performed on USP14-depleted 1205lu cells showed that suppression of USP14 for 6 days altered cell-cycle progression and massively increased cell death as indicated by the appearance of a sub-G1 cell population with reduced DNA content (Fig. 3D). Induction of cell death caused by USP14 knockdown in melanoma cells was further confirmed on USP14-depleted cells stained by Annexin-V-FITC/PI (Fig. 3E) and by immunoblot analysis revealing that USP14 depletion, which reduced Rb phosphorylation, also led to the cleavage of caspase 3 and its substrate PARP, two markers of apoptosis (Fig. 3F). These data confirm that USP14 is an important regulator of melanoma cell survival.
Molecular mechanisms underlying USP14 inhibition
To clarify how UPS14 regulates melanoma cell survival, we first assessed the importance of the caspase-mediated apoptotic process in the cytotoxicity of b-AP15. Flow cytometry analysis of Annexin-V and PI staining performed on cells treated with b-AP15 in the presence or not of the pan-caspase inhibitor QVD-OPh showed that melanoma cell death induced by b-AP15 took place independently of caspase activity (Fig. 4A and B; Supplementary Fig. S5A). USP14 is a DUB predominantly associated with the proteasome, where it cleaves the poly-ubiquitin chains of proteins that are addressed to the proteasome (23, 28). In myeloma cells, USP14 inhibition has been shown to trigger the accumulation of poly-ubiquitinated proteins and an ER stress response (29). We therefore examined how USP14 inhibition affects these events in melanoma cells. Treatment of melanoma cells with b-AP15 induced a rapid accumulation of K48-linked poly-ubiquitinated proteins (Fig. 4C; Supplementary Fig. S6A and S6B), and to a less extent accumulation of K63-linked poly-ubiquitinated proteins (Supplementary Fig. S6C). Targeting USP14 also resulted in the phosphorylation of the stress-related kinases p38 and JNK, and in the upregulation of ER stress–induced proteins Bip/GRP78 and CHOP, as well as in the increased expression of the chaperone HSP70 (Fig. 4C; Supplementary Fig. S6D). Using qRT-PCR analysis, we confirmed that USP14 inhibition in melanoma cells triggered a potent ER stress response as shown by the upregulation of DDIT3/CHOP, HSPA5/GRP78, GADD34, ATF4, and the appearance of the spliced form of XBP1 mRNA (Fig. 4D). Flow cytometry analysis revealed that treatment of melanoma cells with b-AP15 triggered a rapid burst of ROS (Fig. 4E), associated with mitochondrial membrane depolarization (Fig. 4F). Importantly, the ROS scavenger NAC counteracted the anti-melanoma effect of b-AP15 (Supplementary Fig. S5B). Interestingly, compared with bortezomib, b-AP15 did not significantly affect proteasome-associated proteolytic activities (Fig. 4G), nor induce the formation of protein aggresomes (Fig. 4H). Our data indicate that targeting USP14 leads to a caspase-independent cell death associated with accumulation of poly-ubiquitinated proteins and chaperones, ER stress, ROS production, and mitochondrial dysfunction.
Targeting USP14 overcomes resistance to BRAFV600E inhibitors
Acquired resistance to MAPK signaling–targeted drugs remains a clinical challenge in the treatment of advanced metastatic BRAF-mutated melanomas. Therefore, we examined the impact of USP14 inhibition in melanoma cells resistant to BRAFV600E inhibitors (BRAFi). We first used two resistant cell lines that were isolated from parental BRAFV600E melanoma cells following chronic treatment with vemurafenib (isogenic pair M229 and M229R) or dabrafenib (isogenic pair 451lu and 451luR; Fig. 5A; refs. 33, 34). Inhibition of USP14 with b-AP15 potently decreased cell proliferation of BRAFi-resistant cells M229R and 451luR, in a range of concentration that was comparable with their parental counterparts (Fig. 5B). Consistent with this, a DUB trap assay performed on lysates from the two pairs of cells showed that USP14 activity was slightly increased between BRAFi-sensitive and BRAFi-resistant cells (Fig. 5C). We extended these observations to additional isogenic pairs of cells sensitive and resistant to BRAFi through distinct mechanisms (Table 1). Using cell proliferation assays, we calculated the IC50 of b-AP15 on sensitive and BRAFi-resistant cells and we found that b-AP15 blocked proliferation of vemurafenib- or dabrafenib-resistant cells with an efficacy not significantly different to what is observed on the respective parental BRAFi-sensitive cells. Notably, BRAFi-resistant cells were efficiently targeted by b-AP15 regardless of the molecular mechanisms of acquired resistance. Importantly, USP14 inhibition efficiently decreased the viability of A375 melanoma cells with acquired resistance to dual BRAF and ERK inhibition (A375DR cells; Table 1). A colony formation assay carried out on 451lu and 451luR cells further confirmed that b-AP15 could suppress melanoma focus formation independently of acquired resistance to BRAFi (Fig. 5D). Mechanistically, inhibition of USP14 in melanoma cells induced molecular events, including accumulation of K48-linked poly-ubiquitination, decreased Rb phosphorylation, increased p38 phosphorylation, and ER stress response that were indistinguishable between BRAFi-sensitive and BRAFi-resistant cells (Fig. 5E). Our data show that melanoma cell treatment with b-AP15 can overcome resistance to drugs targeting oncogenic BRAF signaling pathway.
Antitumor activity of b-AP15 in a preclinical mouse model of melanoma
To further validate the anti-melanoma activity of b-AP15 observed in vitro, we used a xenograft mouse model of melanoma development in which the BRAFi-resistant cell line 451luR stably expressing the luciferase gene (451LuR-luc+ cells) were injected subcutaneously into nude mice (Fig. 6A). After 3 days, mice were divided into two groups: one group was injected intraperitoneally with b-AP15, while the other group was injected with vehicle alone. Bioluminescence analysis of tumor growth showed a marked decrease in melanoma burden in b-AP15- versus vehicle-treated mice (Fig. 6B). The effect of b-AP15 treatment was already observable after 10 days and maintained up to day 35 of the experiment, as evidenced by BLI and the measurement of the tumor size at endpoint (Fig. 6C). Immunostaining of tumor sections also confirmed the effect of b-AP15 in vivo, with increased apoptosis-related DNA fragmentation in treated mice (Fig. 6D). Importantly, the doses of b-AP15 received by the animals were well tolerated, as no weight loss was observed during the course of the study (Fig. 6E). Our data thus reveal a potent in vivo anti-melanoma activity of b-AP15 and further suggest that targeting USP14 could represent a novel tool to treat melanoma that have acquired resistance to targeted therapy.
Despite the recent introduction of immunotherapies and drugs targeting the MAPK pathway, advanced melanomas remain associated with poor prognosis and as such a clinical challenge. Efforts to identify novel arms to combat metastatic resistant disease are thus warranted. Here we show that the DUB activity of USP14 is specifically elevated during the neoplastic progression of melanocytes to malignant cells. Besides, analysis of public databases first establishes a correlation between USP14 expression and poor outcome in human melanoma. Together, our findings support the notion that USP14 may be pathologically relevant in melanoma.
USP14 is one of the two DUBs that reversibly associate with the 19S regulatory particle of the proteasome (42), to promote the deubiquitination of poly-ubiquitinated substrates addressed to the proteasome (28). The important role of USP14 in regulating proteostasis has been implicated in several types of cancer, including hematologic and solid tumors (23, 24, 29, 30). Our study now identifies USP14 as a key regulator of melanoma cell survival and targeting USP14 with the selective inhibitor b-AP15 and its derivative VLX1570 drastically reduces tumor cell proliferation. Importantly, USP14 inhibition efficiently impairs melanoma cell viability regardless of the mutational status of the oncogenic proteins BRAF and NRAS and of the tumor suppressor genes TP53 and PTEN. These findings could have important clinical consequences for melanoma patients with tumors lacking BRAF mutations. Our data also suggest that elevated and targetable activity of USP14 in melanoma could not be attributed to a particular oncogenic driving event such as mutations affecting BRAF or NRAS. Whether USP14 activity in melanoma could be attributed to the Akt signaling pathway (31) remains to be determined. The collection of cancer cells tested in our study reflects the highly heterogeneous transcriptional and genomic landscape of cutaneous melanoma (2, 6). Notably, b-AP15 and VLX1570 target show similar efficacy on melanoma cells with distinct phenotypes and metastatic potential (3, 4). Given that b-AP15 targets melanoma cells with invasive traits associated with melanoma-initiating activity, and thereby intrinsically resistant to targeted therapy, our results suggest that USP14 inhibition could also be effective in targeting this highly aggressive melanoma cell population.
Another challenge clinicians have to face during the treatment of advanced melanoma is the rapid emergence of drug-resistant melanoma lesions (2, 8). To this regard, we show that selective targeting of USP14 overcomes acquired resistance to BRAFi both in vitro and in vivo. b-AP15 drastically impairs the viability of therapy-resistant cells, in a range of concentration similar to what is required to induce cell death in drug-sensitive cell lines and irrespective of the mechanism of secondary resistance (i.e., NRAS mutation or upregulation of receptor tyrosine kinases). We also demonstrate the anti-melanoma activity of b-AP15 in a preclinical mouse model of BRAFi-resistant tumor growth. This is consistent with previous studies describing antitumor activity of USP14 inhibitors in cancer models (23, 29, 39, 40). Interestingly, inhibition of proteasome-bound DUBs with b-AP15 has been shown to sensitize tumor cells to TRAIL-mediated apoptosis by NK cells and T cells (24). These observations raise the interesting possibility that USP14 inhibition could improve current anti-melanoma immunotherapies that show significant toxicities and patient-selective responses (8).
Previous studies have indicated that DUBs are functionally involved in multiple signaling pathways controlling cell proliferation and migration, as well as cell survival (18, 42). In multiple myeloma, b-AP15 induces a G2–M phase cell-cycle arrest followed by caspase-dependent cell death (29). Consistently, we show that USP14 targeting alters cell-cycle progression and triggers caspase-3 activity and melanoma cell death. However, caspase blockade does not prevent cell death induced by USP14 inhibition, suggesting that USP14 controls melanoma cell viability independently caspase activities. USP14 is predominantly associated with the proteasome, where it deubiquitinates poly-ubiquitinated substrates (28). Consistently, USP14 inhibition triggers in melanoma cells a rapid accumulation of K48-linked poly-ubiquitinated proteins, without significantly affecting the core proteolytic activities of the proteasome, nor the accumulation of toxic protein aggresomes, underlining the selective action of b-AP15. USP14 inhibition also induces a rapid activation of stress-inducible kinases p38 and JNK, accumulation of the chaperone protein HSP70, and a potent ER stress response. Consistent with a previous study (41), the treatment of melanoma cells with b-AP15 also promotes a rapid burst of ROS, followed by mitochondrial dysfunction and an irreversible commitment to cell death. The origin of ROS is currently unknown but our data point toward mitochondria as the actual source of ROS generated downstream of USP14 inhibition. The inhibition of the selenoprotein thioredoxin reductase (TrxR) by b-AP15 (41) represents another possible mechanism that is under investigation. Together, our data indicate that targeting USP14 in melanoma increases oxidative and proteotoxic stress, and ultimately triggers a ROS-dependent and caspase-independent cell death through an unresolved ER stress that overcomes resistance to MAPK-targeting therapies.
Genome sequencing has demonstrated that, compared with other solid tumors, cutaneous melanomas exhibit one of the highest rate of somatic mutations due to carcinogenic ultraviolet light exposure (43). In addition, recent studies have underscored the addiction of melanoma cells to proteostatic processes (44–46). In this context, our findings identifying a role for the proteasome-associated DUB USP14 in melanoma cell survival further document the targetability of aberrant proteostasis in melanoma. A clinical trial investigating VLX1570 for relapsed or refractory multiple myeloma has been suspended due to dose toxicity, indicating that additional drug development is necessary before USP14 inhibitors could enter the clinic. Nevertheless, our study provides a rationale for the evaluation of USP14-targeting drugs as potential therapies to tackle melanoma.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Conception and design: S. Tartare-Deckert, M. Deckert
Development of methodology: R. Didier, A. Mallavialle, M. Tichet, M. Deckert
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Didier, A. Mallavialle, R. Ben Jouira, M.A. Domdom, F. Luciano, M. Ohanna, M. Deckert
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Didier, A. Mallavialle, M. Ohanna, S. Tartare-Deckert, M. Deckert
Writing, review, and/or revision of the manuscript: R. Didier, S. Tartare-Deckert, M. Deckert
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): P. Auberger, S. Tartare-Deckert, M. Deckert
Study supervision: S. Tartare-Deckert, M. Deckert
We thank R.S. Lo, M. Herlyn, J. Villanueva, R. Ballotti, and C. Marine for melanoma cells. We also acknowledge the C3M animal room facility and the C3M imaging facility (Microscopy and Imaging platform Côte d'Azur, MICA). This work was supported by Fondation ARC (grant PJA 20131200347, to M. Deckert), Canceropôle PACA (Emergence 2015, to M. Deckert), Ligue contre le cancer (Equipe labellisée 2016, to S. Tartare-Deckert), the French Government (National Research Agency, ANR) through the “Investments for the Future,” LABEX SIGNALIFE: program reference # ANR-11-LABX-0028–01 (to R. Ben Jouira) and with financial support from ITMO Cancer Aviesan (Alliance Nationale pour les Sciences de la Vie et de la Santé, National Alliance for Life Science and Health) within the framework of the Cancer Plan 2016 (to S. Tartare-Deckert). We also thank financial support by Conseil général des Alpes-Maritimes and Région PACA. R. Didier and R. Ben Jouira were recipients of a doctoral fellowship from Fondation ARC.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.