SPOP Suppresses Hepatocellular Carcinoma Growth and Metastasis by Ubiquitination and Proteasomal Degradation of TRAF6

2.1 Cell Culture

PLC/PRF/5 and HEK293 cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). MHCC-97H, Huh7 and HEK293T cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). DMEM (C11995500BT, Gibco, New York, NY, USA) supplemented with 10% fetal bovine serum (HY-K1006, NTC, Cordoba, Argentina) and 1% penicillin/streptomycin (HY-K1006, MedChem Express, Monmouth Junction, NJ, USA) was used for cell culture. The cells were transfected with Lipofectamine 3000 (L3000150, Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions.

2.2 Antibodies

A list of the antibodies used is provided in Table S1.

2.3 Plasmids

The full-length cDNA of Human TRAF6 and SPOP was amplified using PCR and then ligated into pBudCE4.13-HA vector or pSEB-3Flag vector respectively. TRAF6 and SPOP deletion mutants were generated using two-step PCR amplification. Site-directed mutagenesis was used to generate the SPOP mutants S119N using the wild-type (WT) SPOP-3xFlag plasmid. In addition, the pT7-TRAF6-6 × His (P64553), pCMV-His-UB (P4836), pCMV-UB-K63 × His (P45634) and pCMV-UB-K48 × His (P45628) constructs were purchased from Miaoling Biologica Technology Co. LTD (Wuhan, China). The specific primers are listed in Table S1.

2.4 Immunohistochemistry Analysis

Liver tissue samples were fixed in formalin and processed for immunohistochemical analysis of PCNA and TRAF6. Imaging was conducted using Caseviewer software (3DHISTECH, Budapest, Hungary). The histoscore were evaluated as previously described [30].

2.5 Production of Adenoviruses

The recombinant adenoviruses AdSPOP and AdGFP were generated as previously described [28].

2.6 Clinical Specimens and Mouse Xenograft Assay

During surgical procedures conducted at the First Affiliated Hospital of Chongqing Medical University, tumor tissues and corresponding non-tumorous tissue samples were collected from a cohort of 30 patients. All participants provided informed consent and had not undergone chemotherapy or radiation therapy prior to the surgical intervention.

At 4 weeks of age, BALB/c male nude mice (n = 5 per group) were randomly received 2 × 10⁶ MHCC-97H cells subcutaneously, and tumor size was measured every 3 days thereafter, starting on the sixth day. Tumors from nude mice were extracted, fixed, and stained with PCNA after 30 days.

2.7 Statistics

The differences between two groups were analyzed using the Student's t-test, while differences among multiple groups were evaluated through one-way or two-way ANOVA. Fisher's exact test were performed to analyze the relationship between TRAF6/SPOP expression and clinicopathological features. The data were presented as mean ± SD, and a p-value < 0.05 was defined statistically significant. GraphPad Prism 8 or SPSS v 13 was used for statistical analysis. *p < 0.05, **p < 0.01, ***p < 0.001.

2.8 Others

Other materials and methods are described in the Data S1.

3.1 TRAF6 Is a Novel SPOP-Interacting Protein

Considering that TRAF6 plays a role in NF-κB pathway, we sought to determine whether SPOP exerts tumor-suppressive functions by interacting with TRAF6. We firstly verified the binding between TRAF6 and SPOP using co-immunoprecipitation (Co-IP) assays, which was observed upon ectopic overexpression (Figure 1A,B). This interaction was further substantiated in MHCC-97H cells, where endogenous SPOP and TRAF6 were found to engage in a complex (Figure 1C,D). In addition, the GST pull-down assay confirmed the direct interaction between SPOP and TRAF6 (Figure 1E). Immunofluorescence data further indicated the colocalization of SPOP and TRAF6 in MHCC-97H cells (Figure 1F). To elucidate the subcellular localization of the SPOP and TRAF6 interactions, we conducted Co-IP experiments using extracted cytoplasmic and nuclear proteins. The results demonstrated that the interaction occurs in the cytoplasm (Figure 1G,H). The SPOP-binding consensus (SBC) motif is usually responsible for SPOP binding to its substrates [3]. Therefore, we sought to determine the similar motif in TRAF6 that is required for SPOP-TRAF6 interaction. Despite the absence of a canonical SBC motif, our bioinformatics analysis identified a SBC-like motif (44-NLSSS-48 aa) within TRAF6 sequence (Figure 1I). This motif differs from the classic SBC motif in which the first and second residues were replaced by a polar amino acid, asparagine, and a non-polar amino acid, leucine. We then generated a TRAF6 ΔSBC-like mutant in which the putative motif sequence was deleted. The Co-IP assay showed that TRAF6-ΔSBC-like mutant totally abolished the SPOP-TRAF6 interaction (Figure 1J). Collectively, our findings indicate that SPOP interacts with TRAF6 via the SBC-like motif in TRAF6.

3.2 SPOP Promotes the Ubiquitination and Degradation of TRAF6

Given the established role of SPOP as an E3 ubiquitin ligase in mediating ubiquitination and degradation of its substrates, we first investigated whether SPOP affected TRAF6 protein expression levels. We ectopically expressed HA-tagged TRAF6 in HEK293 cells and observed a significant reduction in TRAF6 protein levels upon co-expression with Flag-SPOP (Figure 2A). The treatment with proteasome inhibitor MG132 effectively rescued TRAF6 levels, implicating the proteasomal pathway in SPOP-mediated TRAF6 degradation. Consistent with these findings, overexpression of SPOP decreased both exogenous and endogenous TRAF6 in a dose-dependent manner, without a concomitant alteration in TRAF6 mRNA level (Figures 2B–D and S1A). Furthermore, we also observed an increased TRAF6 protein levels in SPOP knockout (SPOP-KO) HCC cells (Figure 2E). Upon cycloheximide (CHX) treatment to assess the half-life of TRAF6 protein, we found that SPOP significantly accelerated the degradation of endogenous TRAF6 in MHCC-97H cells (Figure 2F,G). In alignment with these findings, overexpression of SPOP was observed to enhance the ubiquitination of TRAF6 in HCC cells (Figure 2H). It is well-established that K48-linked ubiquitination is generally associated with proteasomal degradation, whereas K63-linked ubiquitination is implicated in non-degradative functions such as cell signaling, inflammatory responses, and cell cycle regulation [31]. Specifically, overexpression of SPOP facilitated the K48-linked polyubiquitination of TRAF6, a modification typically linked to proteasomal degradation (Figure 2I). Conversely, substitution of the lysine at position 48 of ubiquitin with arginine significantly reduced SPOP-mediated ubiquitination of TRAF6, whereas substitution of the lysine at position 63 had no such effect (Figure 2J,K). Collectively, these results suggest that SPOP facilitates the K48-linked polyubiquitination and subsequent degradation of TRAF6 in HCC.

Furthermore, given that the SBC-like motif of TRAF6 is crucial for its interaction with SPOP, we next investigated whether SPOP could promote the ubiquitination and degradation of the TRAF6 ΔSBC-like mutant. The results revealed that the TRAF6 ΔSBC-like mutant abrogated SPOP-mediated degradation and ubiquitination of TRAF6, thereby extending the half-life of TRAF6 protein (Figure S1B–E). These findings collectively demonstrate that SPOP-mediated degradation of TRAF6 is dependent on its SBC-like motif.

3.3 Cancer-Derived SPOP Mutant Lost the Ability to Interact With and Degrade TRAF6

The MATH domain of SPOP is responsible for its substrates recruitment in SPOP (Figure 3A) [21]. We found that deletion of the MATH domain, but not the BTB domain, abolished the interaction between SPOP and TRAF6, and both the MATH and the BTB domain was required for SPOP-mediated TRAF6 ubiquitination and degradation (Figure 3B–D). Subsequently, we investigated the impact of a liver cancer-associated SPOP mutation, S119N, located in MATH domain, on TRAF6 stability [28, 29]. As a result, the S119N mutant was unable to interact with TRAF6 and thereby failed to facilitate the degradation and ubiquitination of TRAF6 proteins in HCC cells (Figure 3E–I). In addition, ectopic expression of wild-type SPOP (SPOP-WT), but not SPOP-S119N mutant, significantly shortened the half-life of TRAF6 (Figure 3J,K). In conclusion, these results indicate that cancer-derived SPOP mutant (S119N in MATH domain) lost the ability to interact with and degrade TRAF6.

3.4 SPOP Negatively Regulates NF-κB Activation Through TRAF6

TRAF6, an upstream regulator of the IKK complex, plays a key role in the NF-κB signaling pathway. We firstly evaluated whether SPOP suppresses the NF-κB pathway. Our findings demonstrated that overexpression of SPOP, in contrast to its knockdown, resulted in a decrease or increase in the levels of phosphorylated TAK1 and p65 in HCC cells, respectively (Figure 4A–D). We next evaluated the impact of SPOP on TRAF6-meditated NF-κB activation and observed that co-expression of SPOP and TRAF6 decreased the levels of phosphorylated TAK1 and p65 when compared to TRAF6 overexpression alone (Figure S2A). Furthermore, we identified an elevation in phosphorylated TAK1 and p65 levels in SPOP-KO cells treated with TRAF6 short hairpin RNA (shRNA) relative to TRAF6 knockdown (Figure S2B). TRAF6 mediated polyubiquitylation of TAK1 has been shown to be required for TAK1 activation [11, 32]. We thus determined whether SPOP could suppress TRAF6-meditated TAK1 ubiquitylation. Results showed that TRAF6 promoted TAK1 ubiquitination and co-transfection of SPOP suppressed TRAF6-meditated TAK1 ubiquitylation in HCC cells (Figure 4E,F). Furthermore, SPOP also suppressed the TRAF6-mediated activation of NF-κB reporter in HCC cells (Figure 4G,H).

NF-κB has been reported to regulate the expression of numerous genes associated with tumor cell survival, proliferation, and invasion [8]. Consequently, we evaluated the expression of NF-κB-dependent oncogenes. Our findings revealed that overexpression of TRAF6 enhanced the expression of NF-κB target genes, including CCND1, IL-8, and VEGFA, in HCC cells. Conversely, this effect is abrogated by co-transfection of SPOP (Figure 4I). In summary, these results suggest that SPOP acts as a negative regulator of TRAF6-mediated NF-κB activation in HCC cells.

3.5 SPOP Negatively Regulates TRAF6-Mediated Cell Proliferation and Invasion

In order to ascertain the involvement of SPOP in TRAF6-mediated HCC cell proliferation and migration, SPOP-KO cells were subjected to TRAF6 shRNA treatment. The results from 5-ethynyl-2′-deoxyuridine (EdU) incorporation assays, clonogenic assays, and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays indicated that knockdown of TRAF6 led to a reduction in cellular proliferation. Conversely, depletion of SPOP resulted in increased cell proliferation and colony-forming capacity in PLC/PRF/5 and MHCC-97H cells (Figures 5A,B, S3A and S4A–C). Moreover, cells with simultaneous depletion of endogenous TRAF6 and SPOP exhibited an enhanced proliferation rate comparable to those with TRAF6 depletion alone (Figures 5A,B, S3A and S4A–C). In keeping with these findings, the wound healing and transwell assays demonstrated that the knockdown of TRAF6 resulted in a decrease in the migration and invasion capabilities of PLC/PRF/5 and MHCC-97H cells, whereas simultaneously knockout of SPOP rewired the effect (Figures 5C, S3B and S4D,E). In addition, simultaneously overexpression of TRAF6 and SPOP reduced the proliferation and invasion capabilities of Huh7 cells when compared to TRAF6 alone (Figure S5A–E). Previous studies have demonstrated that TRAF6 could also reduce apoptosis in cancer cells including HCC cells [20, 33]. Therefore, we further investigated the role of SPOP in TRAF6-mediated reduction of apoptosis in HCC cells. The results revealed that cell apoptosis was increased following treatment with TRAF6 siRNA whereas concurrent knockout of SPOP mitigated this effect in HCC cells (Figures 5D and S4F). Additionally, the simultaneous overexpression of TRAF6 and SPOP led to a higher apoptosis rate in Huh7 cells compared to TRAF6 overexpression alone (Figure S5F).

To extend our in vitro findings, a tumorigenicity model in nude mice injected with MHCC-97H cells was employed to further probe the tumor-suppressive effect of SPOP-mediated ubiquitylation on TRAF6 in vivo. As expected, knockdown of TRAF6 suppressed tumor growth, while knockout of SPOP rescued tumor growth in vivo (Figures 5E–G and S6A). We further validated the expression of downstream molecules of NF-κB, specifically Cyclin D, IL-8, and VEGFA, in xenograft tumors. The results indicated that TRAF6 knockdown reduced the expression levels of Cyclin D, IL-8, and VEGFA, whereas SPOP knockout counteracted this effect (Figure S6B). Altogether, these findings suggest that SPOP can inhibit the proliferation and invasion of HCC cells in vivo and in vitro by targeting TRAF6.

3.6 SPOP Is Correlated With TRAF6 Expression in HCC Tissues

To examine the effect of TRAF6 in HCC patient specimens, we performed the immunohistochemistry (IHC) analysis on six pairs of fixed HCC tissue samples and corresponding non-tumor liver tissues. The results showed that TRAF6 exhibited strong or intermediate staining in HCC tissues compared with non-tumor liver tissues (Figure 6A,B). Subsequently, in order to elucidate the relationship between SPOP and TRAF6 at the protein level in HCC patient specimens, proteins were extracted from 30 pairs of HCC tissue samples and their corresponding non-tumor liver tissues. Western blot analysis revealed a significant upregulation of TRAF6 and a concomitant downregulation of SPOP protein levels in HCC tissues compared to non-tumor liver tissues (Figure 6C–I). Additionally, an inverse correlation was identified between the expression levels of SPOP and TRAF6 (Figure 6J). Furthermore, TRAF6/SPOP expression ratio was positively correlated with the differentiation status and pathological stage of HCC, suggesting potential insights into HCC progression (Table 1). These findings collectively underscore a robust association between the SPOP-TRAF6 axis and human HCC.