Regulation of SOX10 stability via ubiquitination-mediated degradation by Fbxw7α modulates melanoma cell migration.

Dysregulation of SOX10 was reported to be correlated with the progression of multiple cancer types, including melanocytic tumors and tumors of the nervous system. However, the mechanisms by which SOX10 is dysregulated in these tumors are poorly understood. In this study, we report that SOX10 is a direct substrate of Fbxw7α E3 ubiquitin ligase, a tumor suppressor in multiple cancers. Fbxw7α promotes SOX10 ubiquitination-mediated turnover through CPD domain of SOX10. Besides, GSK3β phosphorylates SOX10 at CPD domain and facilitates Fbxw7α-mediated SOX10 degradation. Moreover, SOX10 protein levels were inversely correlated with Fbxw7α in melanoma cells, and modulation of Fbxw7α levels regulated the expression of SOX10 and its downstream gene MIA. More importantly, SOX10 reversed Fbxw7α-mediated suppression of melanoma cell migration. This study provides evidence that the tumor suppressor Fbxw7α is the E3 ubiquitin ligase responsible for the degradation of SOX10, and suggests that reduced Fbxw7α might contribute to the upregulation of SOX10 in melanoma cells.


INTRODUCTION
SRY-related HMG box-containing factor 10 (SOX10) is a transcription factor that belongs to the HMGbox transcription factor family; this protein is initially expressed in premigratory neural crest cells and controls the multipotency, survival, and proliferation of neural crest cells as well as their differentiation into peripheral glial cells and pigment cells at later stages [1]. In addition to its role as a multipotency factor in stem cells, SOX10 has been implicated in the expression of lineage-specific genes in glia and melanocytes [2][3][4]. Homozygous deletion of SOX10 in mice leads to embryonic lethality, whereas SOX10 haploinsufficiency results in a melanocytic phenotype with reduced pigmentation of the belly and limb extremities [5,6]. Previous studies have revealed low frequencies of intragenic mutations of the Sox10 gene in metastatic melanoma, suggesting that SOX10 might be involved in mediating melanoma metastasis [7]. Upregulation of SOX10 protein has been observed in multiple cancer types, including melanocytic tumors and tumors of the nervous system. More recently, a critical role www.impactjournals.com/oncotarget for SOX10 in tumorigenesis and melanoma migration has been demonstrated in cell lines and mouse models [8][9][10][11].
SOX10 expression is tightly regulated at the transcriptional level. Fourteen multiple-species conserved sequences (MCS) were reported to display high levels of evolutionary conservation and variable control of Sox10 expression [12][13][14]. SOXE was identified as binding to MSC4 and MSC7 and thereby enhancing the expression of Sox10. Moreover, four transcriptional factors were found to directly activate Sox10 transcription [1,13]. Autoregulation of Sox10 has been shown in Schwannoma cells [3]. Recently, Sox10 expression was shown to be directly activated in immortalized mammary gland epithelial cells by the TRAP/Drip/Mediator complex, which includes Mediator complex subunit 1 (MED1) and activates gene transcription. MED1 is recruited to the Sox10 promoter at MCS4 and MCS7, and knockdown of MED1 expression completely ablates Sox10 expression in this cell line [15]. The regulation of SOX10 protein at the posttranslational level is less well understood. One study suggested that sumoylation at K55, K246 and K357 of SOX10 by Ubc9 repressed the transcriptional activity of SOX10 [16]. However, the mechanism by which SOX10 protein stability is regulated remains unknown.
In this study, we revealed that SOX10 is an unstable protein, and its stability is controlled by the ubiquitinproteasome proteolytic pathway. Further studies identified Fbxw7α as a potential E3 ubiquitin ligase responsible for SOX10 turnover. Fbxw7α bound to and facilitated the ubiquitination-mediated degradation of SOX10 through phosphodegron. This process is promoted by glycogen synthase kinase 3β (GSK3β)-mediated phosphorylation of SOX10 at the CPD motif. More importantly, we found that Fbxw7α suppresses melanoma cell migration by promoting SOX10 proteolysis. These findings help us to understand the post-translational regulatory mechanism of SOX10 and the underlying clinical significance of the Fbxw7α-SOX10 axis in melanoma.

SOX10 is an unstable protein
To determine whether the SOX10 protein is stable, we assessed the half-life of SOX10 in melanoma cells using the cycloheximide (CHX) chase assay. Aurora-a, a validated unstable protein [31], was used as a positive control. As shown in Figure 1, the SOX10 protein level decreased steadily following protein synthesis inhibition by CHX treatment. The half-life of SOX10 was approximately 4 h. In addition, proteasome inhibitor MG132 treatment induced SOX10 accumulation, suggesting that SOX10 degradation was mediated by ubiquitination.

SOX10 interacts with Fbxw7α
To explore the molecular mechanisms of SOX10 degradation, we sought to identify the E3 ubiquitin ligase responsible for this degradation. Analysis of the amino acid sequence of SOX10 revealed a potential conserved CPD identified in numerous Fbxw7 substrates located between amino acids 235 and 244 of SOX10 ( Figure 2A). Considering that SOX10 is a transcription factor [32] which is usually localized in nucleus, we examined the possibility that SOX10 is a potential substrate of Fbxw7α, the only Fbxw7 isoform localizing in nucleus. Firstly, we tested whether SOX10 interacted with Fbxw7α using coimmunoprecipitation (co-IP). HA-tagged SOX10 was cotransfected with or without Myc-tagged Fbxw7α into 293T cells, and reciprocal co-IP using anti-Myc or anti-HA was performed. As shown in Figure 2B and 2C, the complex containing these two proteins was obviously detected in the cell lysates. Furthermore, the co-localization of Fbxw7α and SOX10 was examined by co-transfected of EGFP-Fbxw7a and dsRed-SOX10 into Hela cells. www.impactjournals.com/oncotarget As shown Figure 2D, Fbxw7α and SOX10 were colocalized in nucleus, and MG132 treatment increased their level in nucleus. We next investigated the crucial domains responsible for their interaction. Mutation of the CPD motif ( Figure 2E) abrogated the interaction between SOX10 and Fbxw7α ( Figure 2G), indicating that the CPD motif was essential for the recognition of SOX10 by Fbxw7α. In addition, deletion of the WD40 domain but not F-box ( Figure 2F) abrogated the interaction between Fbxw7α and SOX10 ( Figure 2H), indicating that Fbxw7α binds to SOX10 through its WD40 domain.

Fbxw7α targets SOX10 for ubiquitination
Fbxw7α is a component of E3 ubiquitin ligase that promotes the degradation of target proteins through ubiquitination. Thus, we used an in vivo ubiquitination assay to test whether Fbxw7α promotes SOX10 ubiquitination. 293T cells transfected with Myc-SOX10 and HA-ubiquitin in the absence or presence of Flag-Fbxw7α were treated with MG132 for 6 h to stabilize the ubiquitinated proteins before lysis. In the absence of ectopic Fbxw7α, SOX10 was weakly ubiquitinated, whereas cotransfection of Fbxw7α increased the ubiquitinated SOX10 level ( Figure 3A). Moreover, deletion of either the F-box or WD40 domain abolished Fbxw7α-induced SOX10 ubiquitination ( Figure 3B). These results indicate that Fbxw7α facilitates the ubiquitination of SOX10.

Fbxw7α facilitates the degradation of SOX10
Based on the observation that Fbxw7α targets SOX10 for ubiquitination, we detected whether Fbxw7α promoted SOX10 turnover. HA-SOX10 was co-transfected with different amounts of Myc-Fbxw7α into 293T cells. Skp2, another F-box containing SCF E3 ubiquitin ligase, was used as a control [33]. Indeed, www.impactjournals.com/oncotarget  Fbxw7α overexpression reduced SOX10 protein levels in a dose-dependent manner, whereas Myc-Skp2 overexpression did not affect SOX10 protein levels ( Figure 4A). Moreover, ectopic expression of HA-Fbxw7α notably reduced the half-life of SOX10 using the CHX chase assay ( Figure 4B), whereas deletion of either the F-box or the WD40 domain abolished Fbxw7α-mediated SOX10 turnover ( Figure 4C). Moreover, mutation of the CPD sequence of SOX10 (SOX10-2A) abrogated its degradation by Fbxw7α ( Figure 4D). Taken together, these results indicate that Fbxw7α is the E3 ubiquitination ligase that mediates SOX10 degradation.

GSK3β is required for the Fbxw7α-mediated degradation of SOX10
Phosphorylation of T/S in the CPD motif of Fbxw7α substrates is required for recognition by Fbxw7α [34]. We sought to determine which phosphokinase is responsible for the phosphorylation of the SOX10 CPD motif. Scansite software analysis revealed that SOX10 CPD is a potential GSK3β phosphorylation motif ( Figure 5A). We next examined the interaction between GSK3β and SOX10 by co-IP. HA-SOX10 was co-transfected with or without Myc-GSK3β into 293T cells, and SOX10 was immunoprecipitated using the anti-HA antibody. GSK3β was copurified with SOX10 only when they were cotransfected ( Figure  5B). Furthermore, mutation of SOX10 CPD (SOX10-2A) abrogated their interaction, indicating that GSK3β interacts specifically with SOX10 and that this interaction depends on the potential GSK3β phosphorylation motif in CPD ( Figure 5C). More importantly, in a in vitro kinase assay, we found that GSK3β directed phosphorylated SOX10, whereas mutation of SOX10 CPD (SOX10-2A) impaired this phosphorylation ( Figure 5D).
We further tested whether GSK3β influenced SOX10 expression. HA-SOX10 was co-transfected with Myc-Fbxw7α into 293T cells with or without GSK3β knockdown, and the SOX10 protein level was monitored by Western blotting. The SOX10 protein level was downregulated upon co-transfecting with Fbxw7α, whereas silencing of GSK3β led to elevation of SOX10 levels compared with the siRNA control ( Figure 5E). In addition, treatment with the GSK3β inhibitors LiCl or AR-A014418 reversed, at least in part, the Fbxw7α-mediated degradation of SOX10 and increased the half-life of SOX10 ( Figure 5F). The half-life of β-Catenin, a well-known GSK3β substrate [35], was also increased upon LiCl or AR-A014418 treatment, indicated that these inhibitors worked well  in the conditions( Figure 5F). Taken together, these results indicate that GSK3β could be the phosphokinase for the phosphorylation of the SOX10 CPD motif, and its kinase activity was required for Fbxw7α-mediated degradation of SOX10.

Fbxw7α regulates endogenous SOX10 in melanoma cells
To further investigate the regulatory relationship between Fbxw7α and SOX10, we first examined the endogenous interaction between Fbxw7α and SOX10 in melanoma cells. A co-IP assay showed that the complex containing the two proteins was obviously detected in melanoma cells using either anti-Fbxw7α or anti-SOX10 antibodies ( Figure 6A). We next examined whether Fbxw7α regulated the endogenous SOX10 level in melanoma cells. Fbxw7α and SOX10 protein levels were detected by Western blotting in a panel of melanoma cells. As shown in Figure 6B, SOX10 was inversely correlated with the Fbxw7α protein levels in melanoma cells. Moreover, Fbxw7α overexpression in MM200 cells downregulated SOX10 expression ( Figure 6C). MIA was reported to be a transcriptional target of SOX10 and responsible for SOX10 mediated melanoma migration [10]. We hence examined whether Fbxw7α regulated MIA level. Indeed, MIA was also downregulated upon Fbxw7α overexpression. Furthermore, co-transfection of GSK3β with Fbxw7α further reduced the protein level of SOX10 compared with Fbxw7α transfected alone ( Figure 6C). By contrast, silencing of Fbxw7α in SK-Mel-Bcl2 cells increased the SOX10 and MIA protein levels ( Figure 6D). Taken together, these results indicate that Fbxw7α regulates the endogenous expression of SOX10 in melanoma cells.

Fbxw7α suppresses melanoma migration through mediation of SOX10 turnover
It has been reported that Fbxw7α suppresses the migration of melanoma cells [36]. We investigated the role of SOX10 in Fbxw7α-mediated migratory inhibition of melanoma cells. SK-Mel-Bcl2 cells were transfected with Fbxw7α siRNAs with or without SOX10 siRNA and were subjected to Transwell and wound-healing assays. The Transwell assay showed that Fbxw7α siRNA transfection dramatically increased the filtered SK-Mel-Bcl2 cells, whereas co-transfection of SOX10 siRNA reduced the filtered cells similar to the negative control ( Figure 7A). Consistently, the wound-healing assay showed that cotransfection of SOX10 siRNA reversed the elevation of the migratory ability of SK-Mel-Bcl2 cells induced by Fbxw7α silencing ( Figure 7B). By contrast, ectopic expression of Fbxw7α suppressed the migration of MM200 cells, whereas the combined transfection of SOX10 reversed Fbxw7αexerted migration suppression effect using the Transwell assay ( Figure 7C) and wound-healing assay ( Figure 7D).

DISCUSSION
In this study, we provide evidence that SOX10 is a direct substrate of SCF Fbxw7α . Fbxw7α interacts with and promotes the ubiquitination-mediated degradation of SOX10 depending on its CPD motif. Moreover, the ubiquitination-dependent degradation of SOX10 by Fbxw7α was enhanced by GSK-3β. Furthermore, Fbxw7αmediated degradation of SOX10 is pathologically relevant, given that SOX10 can reverse the Fbxw7α-mediated migration-suppression effect on melanoma cells.
Sox10 expression is initiated in neural crest cells as they dissociate from the neural tube, and its expression is maintained during neural crest cell migration. Expression transfected with the indicated siRNAs for 48 h, and their migratory ability was tested using a Transwell assay (A) and wound-healing assay (B). C. and D. MM200 cells were transfected with the indicated plasmids for 24 h, and their migratory ability was tested using the Transwell assay (C) and wound-healing assay (D). The results are expressed as the mean ± SD; n = 3, **p < 0.05. continues in the glial and melanocyte lineages, but Sox10 is turned off in many other neural crest cell derivatives [5,32,37]. Both the mRNA and protein level of SOX10 show restricted patterns of tissue-specific expression, suggesting that they undergo dominant regulation at the transcriptional level under physiological conditions. To explore whether SOX10 level was regulated posttranslational, we first accessed the stability of SOX10 using CHX chase assay. Our results indicate that SOX10 was an unstable protein. Besides, the observation that MG132 restored the SOX10 level suggested SOX10 might be degraded through a ubiquitination dependent manner. We hence search for the potential E3 ubiquitin ligase responsible for SOX10 turnover and identified Fbxw7α was the E3 ligase mediated SOX10 degradation. Our results showed that overexpression of Fbxw7α accelerated the SOX10 protein turnover using CHX chase assay. Interestingly, the Fbxw7α protein levels also decrease with CHX treatment. This observation was consistent with previous report that Fbxw7α was unstable due to autoubiquitylation [38]. SOX10 was found to be overexpressed in many cancers, including melanoma, schwannoma, neurofibroma, salivary gland tumors, astrocytoma and glioma [39]. The role of SOX10 in melanoma metastasis has been reported by several studies. Graf et al. examined a panel of melanoma cells and found that SOX10 mRNA amounts varied among melanoma cell lines and did not correlate with progression stage, whereas its protein level was associated with a more invasive or metastatic phenotype, indicating that the SOX10 protein level is regulated posttranslationally [10]. In this study, we found that Fbxw7α promoted SOX10 degradation in melanoma cells. Moreover, the SOX10 protein level was inversely correlated with Fbxw7α in a panel of melanoma cells. These results suggest a major role of the posttranslational regulation of SOX10 by Fbxw7α in melanoma progression.
Fbxw7α has been found to be involved in numerous cellular processes, including cell proliferation, apoptosis, cell cycle and differentiation [20]. Importantly, Fbxw7α is considered a tumor suppressor protein primarily because Fbxw7α targets multiple well-known oncoproteins, including Cyclin E, c-Myc, c-Jun, Mcl-1, and Notch-1 for ubiquitination-mediated destruction [34]. A recent report suggested that Fbxw7α inhibits melanoma migration and may serve as a prognostic marker. The authors found that both Fbw7 protein and mRNA expression was significantly reduced in nine melanoma cell lines compared with normal melanocytes. Moreover, silencing of Fbxw7α results in a remarkable increase in cell migration and stress fiber formation of melanoma cells. However, the authors observed only a subtle change in Fbxw7α substrates such as Myc and Cyclin E upon modulation of Fbxw7α expression Fbxw7αin melanoma cells. These findings suggest that other Fbxw7α substrates mediate the migration of melanoma cells [36]. In the present study, we determined that SOX10 is a novel target of Fbxw7α. Furthermore, a rescue experiment indicates that SOX10 could reverse Fbxw7α-exerted migration inhibition in melanoma cells. These results indicate that Fbxw7α suppresses melanoma metastasis through targeting SOX10 degradation.
In summary, we have shown that SOX10 protein stability was regulated by Fbxw7α-mediated ubiquitination degradation. We also show that Fbxw7α suppressed the SOX10-mediated migration-promoting effect on melanoma cells. Given the frequent downregulation or inactivation of Fbxw7α in melanoma, these findings may help us further understand the roles of the Fbxw7α-SOX10 axis in melanoma progression. Furthermore, the differentiation and development of melanocytes and glia may prove to be another useful model in understanding Fbxw7α-mediated degradation of SOX10, as the SOX10 protein level is attenuated in the differentiation and development of melanocytes and glia.

RNA interference
Fbxw7α siRNAs were purchased from Qiagen (siFbxw7α-1, Qiagen SI03089240) and Abnova (siFbxw7α-2, H00055294-R01). SOX10 siRNAs were purchased from Dharmacon (Smart pool, L-017192). GSK3β was designed according to previously validated oligonucleotides [42] and synthesized by GenePharma (Shanghai, China). Transfection was performed according to the manufacturer's instructions using Lipofectamine™ RNAiMAX transfection reagent (Life Technologies, USA) and 100 nM siRNA. The transfected cells were incubated at 37°C for 48 h in complete medium and were harvested at the indicated time points.

Western blotting and immunoprecipitation
Western blotting and immunoprecipitation were performed as described previously [44]. Briefly, cells were lysed in RIPA buffer [50 mM Tris-HCl at pH 8.0, 2 mM DTT, 5 mM EDTA, 0.5% Nonidet P-40, 100 mM NaCl, 1 mM microcystin, 1 mM sodium orthovanadate, 2 mM phenylmethanesulfonyl fluoride (PMSF), 1 × protease & phosphatase inhibitor cocktail (Thermo Scientific, #1861281)], and clarified lysates were resolved by SDS-PAGE and transferred to PVDF membranes for Western blotting using ECL detection reagents (Advansta, USA; R-03025-D25). Alternatively, clarified supernatants were first incubated with anti-Myc-agarose (Santa Cruz, SC-40AC), anti-FLAG-agarose (Sigma, A2220), or anti-HA-agarose (Sigma, A2095) for 2 h to overnight at 4°C, and the precipitates were washed four times with RIPA buffer. To investigate the interaction between SOX10 and Fbxw7α at the endogenous level, the clarified supernatants were first incubated with anti-Fbxw7α or anti-SOX10 for 2 h at 4°C. Protein A/G-agarose was then added and incubated for 2 h to overnight. Precipitates were washed four times with RIPA buffer and analyzed by Western blotting.

Wound-healing and transwell assays
These procedures were performed as described previously with small modification [45]. Briefly, cells were plated into 6-well-plates and cultured in complete medium supplemented with 20 μM mitomycin C for 24 h. The scraped, acellular area was created with a 200-μL pipette tip. Then the cells were washed with PBS and cultured in DMEM medium with 0.5% FBS and 20 μM mitomycin C. The spread of wound closure was observed after 24 h and imaged under a microscope. Migration assays were performed in modified Boyden chambers with 8-μm pore filter inserts in 24-well plates (BD Transduction, USA). Briefly, 1 × 10 5 cells suspended in serum-free DMEM were added to the upper chamber of the insert in each well of a 24well culture plate. FBS was added to the lower chamber as a chemoattractant at a final concentration of 10%. After 8 h, the nonmigrated cells were gently removed with a cotton swab. The migrated cells in the lower part of the chamber were stained with crystal violet, air dried, and imaged.

Statistical analyses
Statistical analyses were performed using SPSS 16.0 software (SPSS Inc.). The values were expressed as the mean ± standard deviation (SD) of three independent experiments, and the significance of differences between two groups was calculated using two-tailed Student's t-test. P-values less than 0.05 were considered significant.