UBE2N is a K63-specific ubiquitin conjugase linked to various immune disorders and cancer. Here, we demonstrate that UBE2N and its partners UBE2V1 and UBE2V2 are highly expressed in malignant melanoma. Silencing of UBE2N and its partners significantly decreased melanoma cell proliferation and subcutaneous tumor growth. This was accompanied by increased expression of E-cadherin, p16, and MC1R and decreased expression of melanoma malignancy markers including SOX10, Nestin, and ABCB5. Mass spectrometry–based phosphoproteomic analysis revealed that UBE2N loss resulted in distinct alterations to the signaling landscape: MEK/ERK signaling was impaired, FRA1 and SOX10 gene regulators were downregulated, and p53 and p16 tumor suppressors were upregulated. Similar to inhibition of UBE2N and MEK, silencing FRA1 decreased SOX10 expression and cell proliferation. Conversely, exogenous expression of active FRA1 increased pMEK and SOX10 expression, and restored anchorage-independent cell growth of cells with UBE2N loss. Systemic delivery of NSC697923, a small-molecule inhibitor of UBE2N, significantly decreased melanoma xenograft growth. These data indicate that UBE2N is a novel regulator of the MEK/FRA1/SOX10 signaling cascade and is indispensable for malignant melanoma growth. Our findings establish the basis for targeting UBE2N as a potential treatment strategy for melanoma.

Significance: These findings identify ubiquitin conjugase UBE2N and its variant partners as novel regulators of MAPK signaling and potential therapeutic targets in melanoma. Cancer Res; 78(22); 6462–72. ©2018 AACR.

Metastatic melanoma is the most aggressive and difficult-to-treat skin cancer. The incidence of melanoma is on the rise especially among the young population. The NIH SEER program estimated that 87,110 people were diagnosed with melanoma in the United States in 2017, accounting for 5.2% of all new cases of cancer, and that 11% of these patients would succumb to the disease (1). In recent years, immunotherapies and BRAF/MEK oncokinase inhibitors have yielded a high response rate (2–6). However, these treatments fail to produce a long-lasting benefit for the majority of responders due to the rapid development of resistance through cancer cell–intrinsic and -extrinsic mechanisms (7, 8).

The RAS/RAF/MEK/ERK signaling cascade is commonly activated by growth factors and cytokines via an orchestrated cascade of reversible posttranslational modifications, most notably phosphorylation and ubiquitination. In cancer cells, this pathway is often constitutively active as a result of genetic changes. Specifically, BRAF mutation occurs in nearly 70% of cutaneous melanomas and 90% of these mutations are BRAFV600E, which is a potent activator of the downstream MEK/ERK kinases (9). The transformation potency of mutant BRAF is subject to further regulation by ubiquitination (10).

Ubiquitination is a rather complex and multifaceted process (11). Polyubiquitination involves binding of additional ubiquitin monomers to a lysine (K) or methionine (M) residue (e.g., K48, K63, and M1) of the preceding ubiquitin, forming structurally and functionally distinct ubiquitin polymers (Ub). K48-Ub primarily targets proteins to the 26S proteasome complex for degradation, whereas K63-Ub regulates signal transduction and gene expression (12). Ubiquitination typically requires concerted actions of an ubiquitin-activating E1 enzyme, an E2 ubiquitin conjugase, and E3 ubiquitin ligases and it is proteolytically cleaved by deubiquitinases (11). Although E1 is functionally ubiquitous, E2 and E3 enzymes and deubiquitinases are multifaceted. For example, CYLD is a deubiquitinase specific for M1-Ub and K63-Ub and it inhibits inflammation and tumorigenesis (13, 14). In contrast, UBE2N (also called Ubc13) is a K63-Ub–specific E2 enzyme recently characterized as a crucial growth promoter of several human cancers, such as breast cancer, neuroblastoma, B-cell lymphoma, and colon cancer (15–19). UBE2N interacts with a noncatalytic variant UBE2V1 (UeV1) or UBE2V2 (MMS2) to activate NFκB and p38 signaling pathways (15) and DNA repair (20), respectively. However, little is understood about the role of UBE2N in melanoma and the mechanisms underlying its function in cancer.

In this study, we demonstrated that UBE2N and variants are essential for melanoma cell proliferation, survival, and malignant progression. Using unbiased proteomic approach, we revealed a global effect of UBE2N on cell signaling and gene regulation, and identified a MEK/FRA1/SOX10 signaling cascade acting downstream of UBE2N. We also verified FRA1 as a key promoter of pMEK and SOX10 expression and melanoma growth. Finally, we showed the feasibility of pharmacologically interfering with UBE2N function to hinder melanoma xenograft growth in mice.

Cell culture and gene transduction

Cells were cultured at a 37°C incubator supplemented with 5% CO2 and all culture media and supplements were obtained from Thermo Fisher Scientific. A375, A2058, Skmel28, and B16-F10 cells were obtained from ATCC via Duke Cell Culture Facility. DM598, DM733, and DM738 cells were kindly provided by Dr. Hilliard Seigler (Duke University Medical Center, Durham, NC). They were derived from primary biopsies of metastatic melanoma obtained under a Duke University Institutional Review Board (IRB)–approved protocol with written informed consent from patients. These studies were conducted in accordance with U.S. Common Rule. A375, A2058, B16, DM598, DM733, and DM738 were maintained in DMEM with 10% FBS. Cell lines were confirmed to express Mart-1 and Nestin, but no further authentication was performed in this study. A2058, A375, and B16 were negative for all pathogens, except Mycoplasma sp., as screened by PCR (IDEXX BioResearch). Other cell lines were not screened. Normal human melanocytes were isolated from surgically discarded foreskin samples obtained in accordance to an IRB protocol approved by the Duke University (Durham, NC), and cultured in complete melanocyte growth medium 254. The pLKO lentiviral shRNA and CRISPR constructs specific for UBE2N, UBE2V1, UBE2V2, and FRA1 (Supplementary Table S1) were obtained from (Duke Functional Genomics Shared Resource). The retroviral expression vector for the constitutively active FRA1DD mutant was generated as described previously (21). For gene transduction, A375, A2058, and DM598 cells were incubated with the viral media supplemented with 8 μg/mL polybrene overnight in a 37°C incubator supplemented with 5% CO2. Transduced cells were selected with 1 μg/mL puromycin for 1 to 2 weeks.

Cell growth assay and soft-agar colony formation

For cell proliferation assay, cells were seeded onto 96-well dishes at 5,000 cells/well and next day, cells were incubated in quadruples for 48 hours with increasing concentrations of NSC697923 (0, 1, 2, 4, and 8 μmol/L; Selleckchem), cells were incubated with 5 μL 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (20 mg/mL, Sigma-Aldrich) for 2 hours, and media were then replaced with DMSO to dissolve the thiazol crystals. Absorbance was measured using a plate reader (Synergy H1, BioTek). For soft-agar colony formation, 1 mL of 0.5% noble agar was transferred to each well of a 6-well plate to form the base layer. Gene-transduced A375 cells were mixed with 0.35% noble agar, and then seeded in triplicates at 5,000 cells/well onto the solidified base. About 15 minutes later, 1 mL of 10% FBS/DMEM was carefully added to each well. The plates were kept in a 37°C incubator with 5% CO2 and 200 μL of 10% FBS/DMEM fresh media were added to each well every other day. Two weeks later, colonies were stained with nitro blue tetrazolium dye overnight at 37°C. Colonies were imaged and counted using the Olympus microscopic imaging system.

Immunoblotting

Primary human melanocytes, A375, A2058, DM598, DM733, DM738, and B16 cells were grown to about 90% confluence in either the 10%FBS/DMEM medium or the complete melanocyte medium. Protein lysates were collected with RIPA buffer. Transduced cells were reseeded 1 to 2 weeks after selection with puromycin and protein lysates were collected 24 hours later. For Western blotting, proteins (20 μg/sample) were separated by 10%–12% PAGE, transferred onto nitrocellulose membrane, and then blotted with the following antibodies listed in (Supplementary Table S2). The blots were detected with IRDye-conjugated secondary antibodies (Invitrogen) and imaged with the Odyssey Imagining system (Li-COR).

Flow cytometry

A375 cells transduced to express shCon, shUBE2N, shUBE2V1, or shUBE2V2 were fixed with 70% ethanol for 30 minutes at 4°C, treated with 50 μL (100 μg/mL) of ribonuclease solution, and stained with 200 μL of propidium iodide. Cells were analyzed using the BD LSRFortessa Cell Analyzer (BD Biosciences). Cell-cycle analysis was performed using FlowJo software.

Histology and immunostaining

Immunofluorescent staining was performed with frozen tissue sections, as described previously (21). Primary antibodies used for immunofluorescence are listed in (Supplementary Table S2). Samples were counterstained with DAPI (Thermo Fisher Scientific). Images were taken with the Olympus BX41 microscopic imaging system or the Olympus IX73 imaging system.

RT-PCR

Total RNA was isolated from cells using the Qiagen RNAeasy Kit (Qiagen). cDNA was generated by iScript Reverse Transcriptase Supermix and RT-PCR was carried out using the SYBR Green Master Mix (Bio-Rad) in StepOnePlus Real Time Systems (Applied Biosystems). PCR primers spanning the exon–exon junctions were designed specifically for SOX10, FRA1, UBE2N, UBE2V1, UBE2V2, p53, p16, and GAPDH (Supplementary Table S3).

Phosphoproteomic analysis

Phosphoproteomic analysis was performed as described previously (Foster MW and colleagues, JPR 2017), and detailed methods are in Supplementary Methods. Briefly, lysates were prepared in 8 mol/L urea buffer, and following reduction/alkylation, proteins were digested overnight with TPCK-trypsin. Following cleanup, phosphopeptides were enriched using titanium dioxide, and enriched peptides were analyzed using a NanoAcquity UPLC System (Waters) interfaced to a Q Exactive Plus mass spectrometer (Thermo Fisher Scientific). Label-free quantitation using accurate-mass and retention time alignment was performed in Rosetta Elucidator, and database searching was performed using Mascot 2.5 (Matrix Science). Raw data have been deposited to the MassIVE repository (MSV000082208).

Animal studies

Animal studies were performed in compliance with Duke Animal Care and Use Committee. Four- to 6-week-old NSG SCID mice were obtained from Duke Cancer Institute Animal Facility and animals (n = 5/group) were injected subcutaneously in the right and left flank regions with 1.5 × 105 A375 cells expressing shCon, shUBE2N, shUBE2V1, or shUBE2V2 suspended in 100 μL (PBS/Matrigel mixed at 3:1 vol/vol). Tumors were measured biweekly with a caliper for up to 4 weeks, after which, the animals were sacrificed. Tumor tissues were fixed in 10% neutral buffered formalin for histology, flash frozen for protein extraction, or embedded in optimal cutting temperature compound for cryosectioning and immunofluorescence staining. For the pharmacologic study, 8 NSG SCID mice were injected subcutaneously with 1.5 × 105 A375 cells in both flanks. Four mice were randomly selected the next day, and treated every other day for a total of 18 days with either solvent (0.5% DMSO and 30% corn oil) or 5 mg/kg NSC697923. Tumors were measured every 4 days and animals were euthanized for tissue collection as described above.

UBE2N is required for melanoma cell proliferation

To assess the relevance of UBE2N in melanoma, we examined NCBI GEO gene expression datasets, and found that the UBE2N mRNA levels were higher in the metastatic tumors compared with normal skin and benign nevi (Fig. 1A). By immunoblotting, we found that UBE2N, UBE2V1, and UBE2V2 were ubiquitously expressed in melanocytes, but UBE2V2 and UBE2N, particularly the mono-ubiquitinated UBE2N (UBE2N-Ub; ref. 22), exhibited increased expression in the metastatic melanoma cell lines, including A2058, A375, DM598, DM733, DM738, and Skmel28, compared with normal melanocytes. The increased expressions of UBE2N, UBE2V1, and UBE2V2 were especially pronounced in melanoma cells grown with the melanocyte culture media (Fig. 1B). These data indicate that UBE2N and variants are highly expressed in melanoma cells.

Figure 1.

UBE2N is important for melanoma cell proliferation. A, Relative UBE2N mRNA expression levels obtained from NCBI GEO datasets GDS3966 (31 primary and 52 metastatic melanomas) and GDS1375 (7 normal skin, 18 nevi, and 37 malignant melanomas). Graphs represent averages of UBE2N mRNA normalized to that of primary tumor or normal skin ± SD. B, Immunoblotting for UBE2N, UBE2V1, and UBE2V2 in primary human melanocytes and metastatic human melanoma cell lines cultured with either 10% FBS/DMEM or melanocyte culture media. Actin was used as a loading control. C, Confirmation of shRNA-mediated gene silencing via immunoblotting. D, Cell proliferation via MTT assay. Graphs represent average percentages of cell proliferation normalized to control A375 cells at 72 hours after seeding ± SD. P < 0.05. E, Cell-cycle analysis of gene transduced A375 cells via flow cytometry. F, Graph represents average percent of cells in sub-G0, G0, S-, and M-phases ± SD. P values ≤ 0.01 were obtained via unpaired Student t test.

Figure 1.

UBE2N is important for melanoma cell proliferation. A, Relative UBE2N mRNA expression levels obtained from NCBI GEO datasets GDS3966 (31 primary and 52 metastatic melanomas) and GDS1375 (7 normal skin, 18 nevi, and 37 malignant melanomas). Graphs represent averages of UBE2N mRNA normalized to that of primary tumor or normal skin ± SD. B, Immunoblotting for UBE2N, UBE2V1, and UBE2V2 in primary human melanocytes and metastatic human melanoma cell lines cultured with either 10% FBS/DMEM or melanocyte culture media. Actin was used as a loading control. C, Confirmation of shRNA-mediated gene silencing via immunoblotting. D, Cell proliferation via MTT assay. Graphs represent average percentages of cell proliferation normalized to control A375 cells at 72 hours after seeding ± SD. P < 0.05. E, Cell-cycle analysis of gene transduced A375 cells via flow cytometry. F, Graph represents average percent of cells in sub-G0, G0, S-, and M-phases ± SD. P values ≤ 0.01 were obtained via unpaired Student t test.

Close modal

To determine the functional importance of UBE2N, we first performed shRNA-mediated gene silencing of UBE2N, UBE2V1, and UBE2V2 in A375 melanoma cells. The efficiency of gene silencing was as verified by immunoblotting and qRT-PCR (Fig. 1C, Supplementary Fig. S1A–S1C). Consistent with UBE2N being the primary E2 enzyme responsible for K63-Ub, gene silencing of UBE2N and variants markedly reduced K63-Ub of the whole-cell lysates, as shown by immunoblotting (Supplementary Fig. S1D). MTT-based cell growth analysis showed that gene silenced cells grew significantly slower than control cells (Fig. 1D). Furthermore, cell-cycle analysis via flow cytometry revealed that UBE2N loss significantly increased the percentage of cells in the sub-G0 and G0–G1, and inhibited M-phase entry (Fig. 1E and F). To verify the observed growth effects, we performed gene silencing with a second shRNA construct targeting UBE2N and CRISPR constructs targeting UBE2V1 and UBE2V2 in both A375 and A2058 cells. Time-course growth analysis showed that each of these targeting strategies resulted in decreased cell proliferation (Supplementary Fig. S2A and S2B). These data indicate that UBE2N and variants are crucial for melanoma cell proliferation.

UBE2N and variants are required for melanoma growth and progression in vivo

To determine whether UBE2N and its variants are important for melanoma growth in vivo, we performed subcutaneous tumor growth analysis. We found that cells with gene silencing of UBE2N or the variants formed significantly smaller tumors than those of the control cells (Fig. 2A and B). Consistently, these tumors were less proliferative and more apoptotic, as indicated by reduced number of Ki-67–positive cells and increased number of cleaved caspase-3–positive cells, respectively (Fig. 2C and D). In addition, the shUBE2N, shUBE2V1, and the shUBE2V2 tumors showed a markedly increased expression of the epithelial cell marker E-cadherin (Fig. 2E). Interestingly, gene silencing of UBE2N and UBE2V1, but not UBE2V2, reduced expression of the cell adhesion marker β1-Integrin (Fig. 2F).

Figure 2.

Gene silencing of UBE2N inhibits melanoma growth and progression. A, Image of subcutaneous tumors generated in NSG SCID mice with gene-transduced A375 cells. B, Tumor size. Graph depicts average tumor size (n = 10/group) ± SD. P values of less than 0.05 were obtained via unpaired Student t test comparing gene silenced groups to the control group. C–F, Immunofluorescence staining of 4-week-old subcutaneous tumors for Ki-67, cleaved caspase-3, E-cadherin, and β1-Integrin (orange). Nuclei, blue.

Figure 2.

Gene silencing of UBE2N inhibits melanoma growth and progression. A, Image of subcutaneous tumors generated in NSG SCID mice with gene-transduced A375 cells. B, Tumor size. Graph depicts average tumor size (n = 10/group) ± SD. P values of less than 0.05 were obtained via unpaired Student t test comparing gene silenced groups to the control group. C–F, Immunofluorescence staining of 4-week-old subcutaneous tumors for Ki-67, cleaved caspase-3, E-cadherin, and β1-Integrin (orange). Nuclei, blue.

Close modal

Cells with UBE2N loss often appeared bigger and flatter than control cells in culture, which is indicative of differentiation and senescence. By immunostaining, we found that both the melanocyte differentiation marker MC1R and the cell growth inhibitor and senescent marker p16 were significantly increased in the UBE2N, UBE2V1, and UBE2V2 gene–silenced tumors compared with the control tumors (Fig. 3A and B). In contrast, markers previously linked to melanoma malignancy, including SOX10 (23) and Nestin (24), were significantly decreased in these tumors (Fig. 3C and D). These results indicate that UBE2N and its variants are required for melanoma growth and malignant progression.

Figure 3.

UBE2N is required for the suppression of melanoma cell differentiation. A–D, Immunostaining of subcutaneous tumors for MC1R, p16, SOX10, and Nestin (orange). Nuclei, blue. Graphs depict average percentages of MC1R, p16, SOX10, and Nestin-positive cells over total cell population quantified from 5 to 10 images of each group ± SD. P values < 0.01 were obtained via Student t test.

Figure 3.

UBE2N is required for the suppression of melanoma cell differentiation. A–D, Immunostaining of subcutaneous tumors for MC1R, p16, SOX10, and Nestin (orange). Nuclei, blue. Graphs depict average percentages of MC1R, p16, SOX10, and Nestin-positive cells over total cell population quantified from 5 to 10 images of each group ± SD. P values < 0.01 were obtained via Student t test.

Close modal

UBE2N is essential for the expression of a prooncogenic protein landscape

Phosphoproteomics has emerged as an important technique for investigating aberrant signaling in cancer and response to therapy (25, 26). We used mass spectrometry to quantify changes in the phosphoproteomes of A375 cells in response to gene silencing of UBE2N. A total of 7,022 phosphopeptides were quantified (Supplementary Tables S4 and S5), with 2,127 peptides that reached an FDR-corrected P < 0.05; of these 258 were upregulated >5-fold and 509 were downregulated <5-fold in the shUBE2N-expressing versus control cells (Fig. 4A). Contained within these dysregulated phosphopeptides were a number of well-characterized and regulatory sites in a range of progrowth and proinvasion molecules (Fig. 4B). For example, phosphorylation of ERK1/2 (at sites critical for kinase activation) was markedly reduced in shUBE2N-expressing cells while ERK1/2 protein levels were unchanged (Fig. 4C), and this correlated with the reduction of phosphorylation of ERK sensors such as the PXTP site (Thr526) of the transcriptional repressor ERF (Fig. 4B), an Ets family member crucial for cell-cycle progression (27). We also found that the phosphorylation of MEK1/2 (Ser217/221), the upstream activator of ERK1/2, was similarly affected (Fig. 4C). Interestingly, phosphorylation of MEK2 (Thr394), a modification known to be regulated by CDK5 to inhibit the canonical MEK activity (28), was increased in cells with UBE2N loss (Fig. 4B). These effects on the MEK1/2-ERK1/2 pathway were consistent with the antiproliferative effects of UBE2N knockdown.

Figure 4.

UBE2N is essential for the expression of a prooncogenic protein landscape. A, Phosphopeptide expression in A375 cells expressing shUBE2N versus shCon was visualized by volcano plot. Phoshopeptides in red and blue were ≥5-fold higher and ≥5-fold lower, respectively, in shUBE2N cells compared with shCon cells (w/FDR-corrected P < 0.05; unpaired t test). B, Differentially expressed phosphopeptides were queried in the phosphosite database (www.phosphosite.org). The expression values of select phosphosites identified in low-throughput studies (i.e., validated sites with known target kinases and/or regulatory function) were converted to z scores, followed by two-dimensional hierarchical clustering using JMP Pro. C, Protein and phosphoprotein expressions were confirmed by Western blotting. D, mRNA expression was quantified by RT-PCR. Total RNA samples were isolated from A375 cells expressing shCon and shUBE2N. Graphs represent averages of relative mRNA levels of FRA1, SOX10, p53, and p16 normalized to GAPDH ± SD. P values < 0.01 were obtained via unpaired Student t test.

Figure 4.

UBE2N is essential for the expression of a prooncogenic protein landscape. A, Phosphopeptide expression in A375 cells expressing shUBE2N versus shCon was visualized by volcano plot. Phoshopeptides in red and blue were ≥5-fold higher and ≥5-fold lower, respectively, in shUBE2N cells compared with shCon cells (w/FDR-corrected P < 0.05; unpaired t test). B, Differentially expressed phosphopeptides were queried in the phosphosite database (www.phosphosite.org). The expression values of select phosphosites identified in low-throughput studies (i.e., validated sites with known target kinases and/or regulatory function) were converted to z scores, followed by two-dimensional hierarchical clustering using JMP Pro. C, Protein and phosphoprotein expressions were confirmed by Western blotting. D, mRNA expression was quantified by RT-PCR. Total RNA samples were isolated from A375 cells expressing shCon and shUBE2N. Graphs represent averages of relative mRNA levels of FRA1, SOX10, p53, and p16 normalized to GAPDH ± SD. P values < 0.01 were obtained via unpaired Student t test.

Close modal

Although total levels of ERK and MEK1/2 proteins were unchanged, we identified a number of dysregulated phosphoproteins with marked changes in total protein abundance in shUBE2N-expressing cells. These included SOX10 and Fos-related antigen 1 (FRA1), transcription factors important for melanoma progression and therapeutic resistance (Fig. 4C; refs. 23, 29). In addition, we observed an upregulation of p53 and p16 tumor suppressors (Fig. 4C). The relative mRNA levels of FRA1, SOX10, p16, and p53 in shUBE2N versus control cells correlated well with their protein expression, demonstrating that these proteins were transcriptionally regulated (Fig. 4D). To verify the observed molecular changes, we performed immunoblotting with lysates collected from A375 and A2058 cells that had undergone gene targeting with a second shRNA construct for UBE2N and a CRISPR construct for each of the variants. We found that each of these gene-targeting approaches resulted in decreased expression of pMEK, pERK, SOX10, and FRA1 in both A375 and A2058 cells (Supplementary Fig. S2C and S2D). Collectively, these data show that UBE2N regulates an oncogenic proteome at both posttranslational and transcriptional levels.

A375 and A2058 cells express the BRAF(V600E) mutant. We examined whether the observed growth and molecular effects of UBE2N are specific to BRAF-mutant cells. For this, we targeted UBE2N and variants in DM598 cells that harbor an NRAS mutation (30). Interestingly, we found that silencing of UBE2N and variants each reduced cell proliferation of DM598 cells, but the effects were less dramatic than those on A375 and A2058 cells (Supplementary Figs. S2A and S2B and S3A). Immunoblotting showed that pMEK, pERK, SOX10, and FRA1 were not affected by UBE2N loss in DM598 cells (Supplementary Fig. S3B). These data implicate that UBE2N regulation of MEK/ERK signaling might be specific to BRAF-mutant cells.

FRA1 mediates UBE2N regulation of MEK and SOX10, and plays a pivotal role in melanoma growth

To verify that the observed molecular changes were due to reduced UBE2N function, we treated cells with increasing doses of NSC697923, a small-molecule compound that specifically inhibits K63-Ub via disruption of UBE2N interaction with UBE2V1 and UBE2V2 (31). By immunoblotting, we found that NSC697923 decreased FRA1 and SOX10 expression in a dose-dependent manner in both A375 and A2058 cells, as did treatments of the MEK inhibitor trametinib (Supplementary Fig. S4A and S4B), suggesting that UBE2N acts through MEK to regulate FRA1 and SOX10.

FRA1 is an AP1 family gene regulator and promoter of many malignancies including melanoma (29). We asked whether FRA1 loss contributes to the growth defects and other molecular changes observed in shUBE2N cells. To address this question, we first expressed a constitutively active FRA1 mutant (FRA1DD) in A375 cells expressing shUBE2N. By immunoblotting, we found that exogenous FRA1DD expression restored the expressions of pMEK, pERK, and SOX10 that were otherwise downregulated by UBE2N gene silencing or treatment of NSC697923 (Fig. 5A; Supplementary Fig. S4C). ATPase-based cell growth analysis showed that FRA1DD expression restored proliferation of cells with UBE2N-targeted gene silencing (Fig. 5B). In addition, FRA1DD decreased sensitivity of A375 cells to NSC697923 treatment at both high and low serum culture conditions as measured for the 2 and 4 μmol/L doses (Supplementary Fig. S4D and S4E). Moreover, FRA1DD-expressing cells showed increased subcutaneous tumor growth (Supplementary Fig. S4F and S4G).

Figure 5.

FRA1 mediates UBE2N regulation of MEK and SOX10, and plays a pivotal role in melanoma growth. A, Immunoblotting of protein lysates isolated from A375 cells transduced to express shCon, shUBE2N, shUBE2V1, shUBE2V2, and FRA1DD. B, Cell growth analysis via MTT assay. Graph depicts growth curve of A375 cells in triplicates expressing shUBE2N alone or with FRA1DD ± SD. C, Immunoblotting of lysates isolated from A375 and A2058 cells transduced to express shCon or shFRA1. D, Cell growth analysis via MTT assay. Graphs depict time-course growth of triplicates of transduced A375 and A2058 cells expressing shFRA1 normalized to control cells ± SD. E, Representative images of soft-agar colonies of transduced A375 cells. F, Colony counts. Graph represents average number of soft-agar colonies of triplicate dishes ± SD, P values < 0.05 were obtained via unpaired Student t test.

Figure 5.

FRA1 mediates UBE2N regulation of MEK and SOX10, and plays a pivotal role in melanoma growth. A, Immunoblotting of protein lysates isolated from A375 cells transduced to express shCon, shUBE2N, shUBE2V1, shUBE2V2, and FRA1DD. B, Cell growth analysis via MTT assay. Graph depicts growth curve of A375 cells in triplicates expressing shUBE2N alone or with FRA1DD ± SD. C, Immunoblotting of lysates isolated from A375 and A2058 cells transduced to express shCon or shFRA1. D, Cell growth analysis via MTT assay. Graphs depict time-course growth of triplicates of transduced A375 and A2058 cells expressing shFRA1 normalized to control cells ± SD. E, Representative images of soft-agar colonies of transduced A375 cells. F, Colony counts. Graph represents average number of soft-agar colonies of triplicate dishes ± SD, P values < 0.05 were obtained via unpaired Student t test.

Close modal

Next, we performed FRA1-targeted gene silencing with lentiviral shRNA in A375 and A2058 cells. We found that FRA1 gene silencing decreased pMEK, pERK, and SOX10 expressions (Fig. 5C), and inhibited proliferation of both cell lines (Fig. 5D). Furthermore, soft-agar colony formation assay showed that gene silencing of FRA1, UBE2N, UBE2V1, or UBE2V2 each resulted in a significant decrease of colony formation, whereas FRA1DD increased it (Fig. 5E and F). These data indicated that FRA1 is an important mediator of UBE2N regulation of MEK/SOX10 signaling, and plays a pivotal role in melanoma growth.

Pharmacologic inhibition of UBE2N mitigates melanoma growth in vivo

Consistent with gene silencing effects, treatment of NSC697923 inhibited the proliferation of A375, A2058, and B16 melanoma cell lines in a dose–response manner, and sensitized cells to death under serum-starved conditions (Fig. 6A–C). In addition, we observed that NSC697923 significantly slowed cell migration, as demonstrated by scratch wound assay (Supplementary Fig. S5A and S5B).

Figure 6.

Systemic treatment of NSC697923 inhibits melanoma xenograft growth and decreases melanoma tumor growth and malignancy. A, Cell growth analysis via MTT assay. A375, A2058, and B16 were treated with varying concentrations of NSC697923 in the presence of 1% or 10% FBS. Graphs represent averages of tetrad wells ± SD. P values < 0.05 were obtained via unpaired Student t test. B, Subcutaneous tumor growth kinetics. A375 cells (1.5 × 105/injection) were injected into both flanks of NSG mice. Animals (n = 4/group) were then treated every other day with intraperitoneal dose of 100-μL solvent (5% DMSO in 30% corn oil) or NSC697923 (5 mg/kg body weight). C, Images of the subcutaneous tumors collected 22 days after injection. D, Immunoblotting of protein lysates isolated from subcutaneous tumors for K63-Ub and Actin. E–J, Immunostaining of the frozen sections of the subcutaneous tumors for Ki-67, cleaved caspase-3, SOX10, Nestin, ABCB5, and p16 (orange). Nuclei, blue. Graphs depict quantification of relative percentages or intensities of cells positively stained for the above markers. P values < 0.05 were obtained via unpaired Student t test.

Figure 6.

Systemic treatment of NSC697923 inhibits melanoma xenograft growth and decreases melanoma tumor growth and malignancy. A, Cell growth analysis via MTT assay. A375, A2058, and B16 were treated with varying concentrations of NSC697923 in the presence of 1% or 10% FBS. Graphs represent averages of tetrad wells ± SD. P values < 0.05 were obtained via unpaired Student t test. B, Subcutaneous tumor growth kinetics. A375 cells (1.5 × 105/injection) were injected into both flanks of NSG mice. Animals (n = 4/group) were then treated every other day with intraperitoneal dose of 100-μL solvent (5% DMSO in 30% corn oil) or NSC697923 (5 mg/kg body weight). C, Images of the subcutaneous tumors collected 22 days after injection. D, Immunoblotting of protein lysates isolated from subcutaneous tumors for K63-Ub and Actin. E–J, Immunostaining of the frozen sections of the subcutaneous tumors for Ki-67, cleaved caspase-3, SOX10, Nestin, ABCB5, and p16 (orange). Nuclei, blue. Graphs depict quantification of relative percentages or intensities of cells positively stained for the above markers. P values < 0.05 were obtained via unpaired Student t test.

Close modal

To establish the utility of NSC696723 for in vivo applications, we examined its effect on the A375 xenograft model. We found that intraperitoneal injections of NSC697923 (5 mg/kg) administered every other day significantly reduced subcutaneous tumor growth of A375 cells in immunodeficient NSG mice (Fig. 6B and C). Immunoblotting of tumor protein lysates confirmed that total K63-Ub was reduced by about 2-fold in tissues treated with NSC697923 (Fig. 6D). By immunostaining, we found that tumors treated with NSC697923 were less proliferative and more apoptotic than the control tumors, as indicated by the reduced number of Ki-67 cells (Fig. 6E) and the increased number of cleaved caspase-3–positive cells, respectively (Fig. 6F). In addition, we found that NSC697923-treated tumors expressed markedly reduced levels of SOX10, Nestin, and ABCB5 (Fig. 6G–I). In contrast, p16 was increased in the drug-treated tissues (Fig. 6J). During the course of treatment, none of the treated animals showed any superficial signs of toxicity or significant weight loss (Supplementary Fig. S6A–S6C). These results indicate that systemic delivery of NSC697923 impedes melanoma growth by decreasing the expression of melanoma malignancy markers and increasing the expression of tumor suppressors.

Our study demonstrates a critical role of UBE2N and its variant partners in melanoma growth and progression. We characterized UBE2N as a negative regulator of tumor suppressors and a positive regulator of oncogenic proteins. Specifically, our data support a working model where FRA1 is a crucial effector molecule acting downstream of UBE2N to maintain the MEK/FRA1/SOX10 signaling cascade and melanoma malignancy (Fig. 7). Most importantly, we established the preclinical feasibility of applying a small-molecule inhibitor of UBE2N to attenuate melanoma growth in vivo. Our findings underscore UBE2N as a new promising therapeutic target for melanoma. Given that UBE2N is relevant to many other cancers (15–19), the mechanistic insights obtained from this study have broad implications.

Figure 7.

Working model for UBE2N function in melanoma. UBE2N acts in part through a MEK/FRA1/SOX10 signaling cascade to promote melanoma growth, survival, and malignant progression.

Figure 7.

Working model for UBE2N function in melanoma. UBE2N acts in part through a MEK/FRA1/SOX10 signaling cascade to promote melanoma growth, survival, and malignant progression.

Close modal

Previous studies have linked UBE2N to NFκB, p38, and p53 signaling pathways in several cancer models (15–19). Our studies have for the first time identified a novel MEK/FRA1/SOX10 signaling cascade downstream of UBE2N. It is rather intriguing that UBE2N regulation appears to be specific to BRAF but not NRAS-mutant cells. BRAF(V600E) is subject to K63-Ub, which is required for BRAF(V600E)-driven transformation (10). Although our attempts to detect BRAF-K63-Ub were not successful, we observed that UBE2N loss increased MEK2 phosphorylation at a noncanonical Thr394 residue, a modification previously shown to be responsible for CDK5-mediated MEK/ERK inactivation and control of insulin resistance (28). Furthermore, studies are needed to determine whether this is the main mechanism underlying MEK inactivation by UBE2N loss. It is also interesting to observe that knockdown of UBE2V2 appears to induce a more potent effect on cell proliferation than that of other subunits. Like any other gene-targeting approach, there is a risk of off-target effects. RT-PCR and immunoblotting data show that UBE2N and UBE2V1 are highly expressed in shUBE2V2 cells, suggesting that nonspecific targeting of UBE2N and UBE2V1 is unlikely the main cause of the dramatic phenotype. Although the high efficiency of UBE2V2 knockdown and other unknown nonspecific effects might contribute to the observed phenotype, it is conceivable that UBE2V2 has UBE2N-independent functions, as were recently reported for UBE2N (32, 33). It is also clear that UBE2V2 has MEK-independent functions (20). This is shown by our data demonstrating that knockdown of UBE2V2 significantly decreased cell proliferation of DM598 cells, but it has a negligible effect on MEK activation. Furthermore, studies are required to better understand the distinct mechanisms mediating the functions of each UBE2N subunit and how these proteins crosstalk with BRAF oncogene.

As an AP1 family transcription factor, FRA1 regulates expression of genes involved in cell-cycle progression, tumor cell migration, and angiogenesis (29). In addition, FRA1 regulates membrane lipid synthesis independent of its transcriptional activity and subsequently AKT activation (21, 34). FRA1-targeted gene silencing decreased MEK/ERK activation, and MEK inhibition decreased FRA1 expression, supporting the existence of a positive feedback signaling loop. FRA1 silencing also slightly downregulated the expression of UBE2V2 (Fig. 5C). FRA1 is known to be induced by MEK (35) and interestingly, in turn activates MEK/ERK signaling through AP1 target genes such as HBEGF, thereby maintaining a positive feedback signaling loop (36). Because the effects of UBE2N loss on MEK and FRA1 both appear to be specific to BRAF-mutant cells, it is conceivable that they are regulated through a BRAF-related common target of UBE2N.

SOX10 is a neuronal cell marker expressed at a basal level in normal human melanocytes, but its expression is significantly upregulated in malignant melanomas (37). SOX10 regulates the expression of Nestin, another neural progenitor cell marker associated with melanoma growth and migration (38). ABCB5 belongs to a family of membrane transporters that regulate membrane potential and chemoresistance (39). The downregulation of SOX10, Nestin, and ABCB5 by genetic and/or pharmacologic inhibition of UBE2N supports that UBE2N is essential for malignant growth of melanoma cells. Furthermore, we established that in melanoma, SOX10 is regulated by UBE2N via FRA1. In line with our findings, UBE2V1 and UBE2V2 are overexpressed in human embryonic stem cells and cancer cell lines compared with their normal adult counterparts (40). It is highly conceivable that UBE2N and variants play a role in cancer stem cell phenotype.

Finally, UBE2N may also be involved in modulation of a tumor microenvironment that favors tumor progression and resistance to therapy. In this regard, UBE2N loss altered expression of a plethora of molecules relevant to extracellular matrix and immune function (Supplementary Table S5). Furthermore, studies are needed to understand how UBE2N regulates the metabolic landscape of the tumor microenvironment, which by itself provides a goldmine of opportunities for new cancer therapeutics (41). UBE2N also inhibits the conversion of regulatory T cells into effector T cells (42), suggesting that treatment of UBE2N inhibitors such as NSC697923 could improve the antitumor immune response stimulated by immune checkpoint inhibitors.

No potential conflicts of interest were disclosed.

Conception and design: A. Dikshit, J.Y. Zhang

Development of methodology: A. Dikshit, J.Y. Zhang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Dikshit, S. Degan, J. Hwang, M.W. Foster, J.Y. Zhang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Dikshit, S. Degan, J. Hwang, M.W. Foster, J.Y. Zhang

Writing, review, and/or revision of the manuscript: A. Dikshit, M.W. Foster, J.Y. Zhang

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Dikshit, Y.J. Jin, C.-Y. Li, J.Y. Zhang

Study supervision: J.Y. Zhang

This work was funded by NCI (RCA188619 to J.Y. Zhang) and Department of Dermatology of Duke School of Medicine. We thank So Young Kim (Duke Functional Genomics Shared Resource) for providing the CRISPR and shRNA constructs and Arthur Moseley (Duke Proteomic Core) for guidance in proteome analysis.

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.

1.
Howlader
N
,
Krapcho
M
,
Miller
D
,
Bishop
K
,
Kosary
CL
,
Yu
M
, et al
SEER Cancer Statistics Review, 1975–2014 [monograph on the Internet]
.
Bethesda, MD
:
NCI
; 
2016
.
Available from
: https://seer.cancer.gov/archive/csr/1975_2014/.
2.
Hodi
FS
,
O'Day
SJ
,
McDermott
DF
,
Weber
RW
,
Sosman
JA
,
Haanen
JB
, et al
Improved survival with ipilimumab in patients with metastatic melanoma
.
N Engl J Med
2010
;
363
:
711
23
.
3.
Robert
C
,
Thomas
L
,
Bondarenko
I
,
O'Day
S
,
Weber
J
,
Garbe
C
, et al
Ipilimumab plus dacarbazine for previously untreated metastatic melanoma
.
N Engl J Med
2011
;
364
:
2517
26
.
4.
Luke
JJ
,
Ott
PA
. 
PD-1 pathway inhibitors: the next generation of immunotherapy for advanced melanoma
.
Oncotarget
2015
;
6
:
3479
92
.
5.
Chapman
PB
,
Hauschild
A
,
Robert
C
,
Haanen
JB
,
Ascierto
P
,
Larkin
J
, et al
Improved survival with vemurafenib in melanoma with BRAF V600E mutation
.
N Engl J Med
2011
;
364
:
2507
16
.
6.
Flaherty
KT
,
Robert
C
,
Hersey
P
,
Nathan
P
,
Garbe
C
,
Milhem
M
, et al
Improved survival with MEK inhibition in BRAF-mutated melanoma
.
N Engl J Med
2012
;
367
:
107
14
.
7.
Zhu
J
,
Powis de Tenbossche
CG
,
Cane
S
,
Colau
D
,
van Baren
N
,
Lurquin
C
, et al
Resistance to cancer immunotherapy mediated by apoptosis of tumor-infiltrating lymphocytes
.
Nat Commun
2017
;
8
:
1404
.
8.
O'Donnell
JS
,
Long
GV
,
Scolyer
RA
,
Teng
MW
,
Smyth
MJ
. 
Resistance to PD1/PDL1 checkpoint inhibition
.
Cancer Treat Rev
2017
;
52
:
71
81
.
9.
Pollock
PM
,
Harper
UL
,
Hansen
KS
,
Yudt
LM
,
Stark
M
,
Robbins
CM
, et al
High frequency of BRAF mutations in nevi
.
Nat Genet
2003
;
33
:
19
20
.
10.
An
L
,
Jia
W
,
Yu
Y
,
Zou
N
,
Liang
L
,
Zhao
Y
, et al
Lys63-linked polyubiquitination of BRAF at lysine 578 is required for BRAF-mediated signaling
.
Sci Rep
2013
;
3
:
2344
.
11.
Yau
R
,
Rape
M
. 
The increasing complexity of the ubiquitin code
.
Nat Cell Biol
2016
;
18
:
579
86
.
12.
Gallo
LH
,
Ko
J
,
Donoghue
DJ
. 
The importance of regulatory ubiquitination in cancer and metastasis
.
Cell Cycle
2017
;
16
:
634
48
.
13.
Mathis
BJ
,
Lai
Y
,
Qu
C
,
Janicki
JS
,
Cui
T
. 
CYLD-mediated signaling and diseases
.
Curr Drug Targets
2015
;
16
:
284
94
.
14.
Ke
H
,
Augustine
CK
,
Gandham
VD
,
Jin
JY
,
Tyler
DS
,
Akiyama
SK
, et al
CYLD inhibits melanoma growth and progression through suppression of the JNK/AP-1 and beta1-integrin signaling pathways
.
J Invest Dermatol
2013
;
133
:
221
9
.
15.
Wu
Z
,
Shen
S
,
Zhang
Z
,
Zhang
W
,
Xiao
W
. 
Ubiquitin-conjugating enzyme complex Uev1A-Ubc13 promotes breast cancer metastasis through nuclear factor-кB mediated matrix metalloproteinase-1 gene regulation
.
Breast Cancer Res
2014
;
16
:
R75
.
16.
Pulvino
M
,
Liang
Y
,
Oleksyn
D
,
DeRan
M
,
Van Pelt
E
,
Shapiro
J
, et al
Inhibition of proliferation and survival of diffuse large B-cell lymphoma cells by a small-molecule inhibitor of the ubiquitin-conjugating enzyme Ubc13-Uev1A
.
Blood
2012
;
120
:
1668
77
.
17.
Cheng
J
,
Fan
YH
,
Xu
X
,
Zhang
H
,
Dou
J
,
Tang
Y
, et al
A small-molecule inhibitor of UBE2N induces neuroblastoma cell death via activation of p53 and JNK pathways
.
Cell Death Dis
2014
;
5
:
e1079
.
18.
Gombodorj
N
,
Yokobori
T
,
Yoshiyama
S
,
Kawabata-Iwakawa
R
,
Rokudai
S
,
Horikoshi
I
, et al
Inhibition of ubiquitin-conjugating enzyme E2 may activate the degradation of hypoxia-inducible factors and, thus, overcome cellular resistance to radiation in colorectal cancer
.
Anticancer Res
2017
;
37
:
2425
36
.
19.
Wu
X
,
Zhang
W
,
Font-Burgada
J
,
Palmer
T
,
Hamil
AS
,
Biswas
SK
, et al
Ubiquitin-conjugating enzyme Ubc13 controls breast cancer metastasis through a TAK1-p38 MAP kinase cascade
.
Proc Natl Acad Sci U S A
2014
;
111
:
13870
5
.
20.
Andersen
PL
,
Zhou
H
,
Pastushok
L
,
Moraes
T
,
McKenna
S
,
Ziola
B
, et al
Distinct regulation of Ubc13 functions by the two ubiquitin-conjugating enzyme variants Mms2 and Uev1A
.
J Cell Biol
2005
;
170
:
745
55
.
21.
Zhang
X
,
Wu
J
,
Luo
S
,
Lechler
T
,
Zhang
JY
. 
FRA1 promotes squamous cell carcinoma growth and metastasis through distinct AKT and c-Jun dependent mechanisms
.
Oncotarget
2016
;
7
:
34371
83
.
22.
Bekes
M
,
Okamoto
K
,
Crist
SB
,
Jones
MJ
,
Chapman
JR
,
Brasher
BB
, et al
DUB-resistant ubiquitin to survey ubiquitination switches in mammalian cells
.
Cell Rep
2013
;
5
:
826
38
.
23.
Graf
SA
,
Busch
C
,
Bosserhoff
A-K
,
Besch
R
,
Berking
C
. 
SOX10 promotes melanoma cell invasion by regulating melanoma inhibitory activity
.
J Invest Dermatol
2014
;
134
:
2212
20
.
24.
Akiyama
M
,
Matsuda
Y
,
Ishiwata
T
,
Naito
Z
,
Kawana
S
. 
Inhibition of the stem cell marker nestin reduces tumor growth and invasion of malignant melanoma
.
J Invest Dermatol
2013
;
133
:
1384
7
.
25.
Basken
J
,
Stuart
SA
,
Kavran
AJ
,
Lee
T
,
Ebmeier
CC
,
Old
W
, et al
Specificity of phosphorylation responses to MAP kinase pathway inhibitors in melanoma cells
.
Mol Cell Proteomics
2017
.
26.
Stuart
SA
,
Houel
S
,
Lee
T
,
Wang
N
,
Old
WM
,
Ahn
NG
. 
A phosphoproteomic comparison of B-RAFV600E and MKK1/2 inhibitors in melanoma cells
.
Mol Cell Proteomics
2015
;
14
:
1599
615
.
27.
Le Gallic
L
,
Sgouras
D
,
Beal
G
 Jr
,
Mavrothalassitis
G
. 
Transcriptional repressor ERF is a Ras/mitogen-activated protein kinase target that regulates cellular proliferation
.
Mol Cell Biol
1999
;
19
:
4121
33
.
28.
Banks
AS
,
McAllister
FE
,
Camporez
JP
,
Zushin
PJ
,
Jurczak
MJ
,
Laznik-Bogoslavski
D
, et al
An ERK/Cdk5 axis controls the diabetogenic actions of PPARgamma
.
Nature
2015
;
517
:
391
5
.
29.
Maurus
K
,
Hufnagel
A
,
Geiger
F
,
Graf
S
,
Berking
C
,
Heinemann
A
, et al
The AP-1 transcription factor FOSL1 causes melanocyte reprogramming and transformation
.
Oncogene
2017
;
36
:
5110
21
.
30.
Augustine
CK
,
Toshimitsu
H
,
Jung
S-H
,
Zipfel
PA
,
Yoo
JS
,
Yoshimoto
Y
, et al
Sorafenib, a multikinase inhibitor, enhances the response of melanoma to regional chemotherapy
.
Mol Cancer Ther
2010
;
9
:
2090
101
.
31.
Hodge
CD
,
Edwards
RA
,
Markin
CJ
,
McDonald
D
,
Pulvino
M
,
Huen
MS
, et al
Covalent inhibition of Ubc13 affects ubiquitin signaling and reveals active site elements important for targeting
.
ACS Chem Biol
2015
;
10
:
1718
28
.
32.
Huen
MSY
,
Huang
J
,
Yuan
J
,
Yamamoto
M
,
Akira
S
,
Ashley
C
, et al
Noncanonical E2 variant-independent function of UBC13 in promoting checkpoint protein assembly
.
Mol Cell Biol
2008
;
28
:
6104
12
.
33.
Brun
J
,
Chiu
R
,
Lockhart
K
,
Xiao
W
,
Wouters
BG
,
Gray
DA
. 
hMMS2 serves a redundant role in human PCNA polyubiquitination
.
BMC Mol Biol
2008
;
9
:
24
.
34.
Motrich
RD
,
Castro
GM
,
Caputto
BL
. 
Old players with a newly defined function: Fra-1 and c-Fos support growth of human malignant breast tumors by activating membrane biogenesis at the cytoplasm
.
PLoS One
2013
;
8
:
e53211
.
35.
Verde
P
,
Casalino
L
,
Talotta
F
,
Yaniv
M
,
Weitzman
JB
. 
Deciphering AP-1 function in tumorigenesis: fra-ternizing on target promoters
.
Cell Cycle
2007
;
6
:
2633
9
.
36.
Mullenbrock
S
,
Shah
J
,
Cooper
GM
. 
Global expression analysis identified a preferentially nerve growth factor-induced transcriptional program regulated by sustained mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) and AP-1 protein activation during PC12 cell differentiation
.
J Biol Chem
2011
;
286
:
45131
45
.
37.
Mohamed
A
,
Gonzalez
RS
,
Lawson
D
,
Wang
J
,
Cohen
C
. 
SOX10 expression in malignant melanoma, carcinoma, and normal tissues
.
App Immunohistochem Mol Morphol
2013
;
21
:
506
10
.
38.
Klein
WM
,
Wu
BP
,
Zhao
S
,
Wu
H
,
Klein-Szanto
AJ
,
Tahan
SR
. 
Increased expression of stem cell markers in malignant melanoma
.
Mod Pathol
2007
;
20
:
102
7
.
39.
Frank
NY
,
Margaryan
A
,
Huang
Y
,
Schatton
T
,
Waaga-Gasser
AM
,
Gasser
M
, et al
ABCB5-mediated doxorubicin transport and chemoresistance in human malignant melanoma
.
Cancer Res
2005
;
65
:
4320
33
.
40.
Ma
L
,
Broomfield
S
,
Lavery
C
,
Lin
SL
,
Xiao
W
,
Bacchetti
S
. 
Up-regulation of CIR1/CROC1 expression upon cell immortalization and in tumor-derived human cell lines
.
Oncogene
1998
;
17
:
1321
6
.
41.
Sugiura
A
,
Rathmell
JC
. 
Metabolic barriers to T cell function in tumors
.
J Immunol
2018
;
200
:
400
7
.
42.
Chang
JH
,
Xiao
Y
,
Hu
H
,
Jin
J
,
Yu
J
,
Zhou
X
, et al
Ubc13 maintains the suppressive function of regulatory T cells and prevents their conversion into effector-like T cells
.
Nat Immunol
2012
;
13
:
481
90
.