This study collected human HCC tissue samples and paired adjacent tissue samples from 80 patients with HCC in the First Affiliated Hospital of Xi’an Jiaotong University. The included patients had precise pathological diagnoses and had not been treated before surgery. All the collected tissue samples were well were formalin-fixed and paraffin-embedded. Written informed consent was obtained from all patients, and the Ethics Committee of the 1st Affiliated Hospital of Xi’an Jiaotong University approved this study. During 3 years of follow-up, the time between primary surgery and death was the overall survival time.

Hep3B, HepG2, HCCLM3 and Huh7 cells were provided by Stem Cell Bank, Chinese Academy of Sciences (Shanghai, China). Human liver cell line MIHA was purchased from bnbio (Beijing, China). Cell lines were cultured in DMEM (Gibco, Thermo Fisher Scientific, Waltham, MA, United States) added with 10% fetal bovine serum (Gibco), penicillin (100 U/ml, Gibco) and streptomycin (100 mg/ml, Gibco). The culture condition in the incubator was a moist environment with 5% CO₂ at 37°C.

LKO.1-puro lentiviral vectors encoding small hairpin RNA (shRNA) against HIF-1α (sh1α), HIF-2α (sh2α), RNF146 (shRNF#1 and shRNF#2), PTEN (shPTEN) and non-targeting shRNA (NTC) were provided by GeneChem (Shanghai, China). All lentiviral shuttle vectors were transfected into HEK293T cells for packaging as previously described (Zeng et al., 2021). RNF146 cDNA was sub-cloned into pcDNA3.1 (Invitrogen, Carlsbad, CA, United States) to produce pcDNA3.1/RNF146 (RNF-OE). Lipofectamine 3000 (Thermo Fisher Scientific) was applied for plasmid transfection in HCC cells.

HCC cells were subjected to RNA isolation following the instruction of the Trizol reagent (Invitrogen) and total RNA was reversely transcribed into cDNA using PrimeScript™ RT Master Mix (Takara, Shiga, Japan). The PCR amplification was performed on ABI HT9600 (Applied Biosystems, Foster City, CA, United States) in accordance with the instruction of SYBR Green PCR Master Mix (Takara). The RNF146 mRNA level was normalized to GAPDH using the 2^−ΔΔCT method. RNF146 forward primer: 5′-TGT AAG CAC GTT TTC TGC TAT CT-3′; reverse: 5′-AAT CCT CGG GAA TTT CTT GTC G-3′. GAPDH forward primer: 5′-CTG GGC TAC ACT GAG CAC C-3′; reverse: 5′-AAG TGG TCG TTG AGG GCA ATG-3′.

Total proteins were obtained from HCC cells using RIPA buffer (Beyotime, Shanghai, China) and quantified by Enhanced BCA Protein Assay Kit (Beyotime). Then, 20 μg proteins were undergoing SDS-PAGE and transferred to the PVDF membrane (Millipore, Billerica, MA, United States). After blocking, the membrane was incubated overnight with primary antibodies at 4°C. The next day, secondary antibody incubation was performed at room temperature for 1–2 h. Enhanced chemiluminescence (ECL) reagent (Millipore) was used for luminous reaction and western blot was photographed by Amersham Imager 680 (GE Healthcare Life Sciences, Pittsburgh, PA, United States). The used primary antibodies: HIF-1α (ab1, Abcam, Cambridge, MA, United States), HIF-2α (ab199, Abcam), RNF146 (ab201212, Abcam), β-tubulin (10094-1-AP, Proteintech, Wuhan, China), PTEN (ab267787, Abcam). p-AKT (Ser473, 66444-1-Ig, Proteintech), AKT (60203-2-Ig, Proteintech), p-mTOR (Ser2448, 67778-1-Ig, Proteintech), mTOR (66888-1-Ig, Proteintech), Ubiquitin (ab7254, Abcam).

The embedded-paraffin tissues were sectioned with a thickness of approximately 4–5 μm. Next, paraffin sections were deparaffinized in xylene and rehydrated with ethanol, and used 0.3% hydrogen peroxide to block endogenous peroxidase activity. Finally, the sections were incubated with RNF146 antibody (ab201212, Abcam) and a specific secondary antibody. The intensity score and positive rate score were previously described (Shen et al., 2018). The IHC score (intensity score × positive rate score) ≥4 was defined as high expression.

The immunoprecipitation and ubiquitination assays were performed as previously described (Shi et al., 2021). In brief, the cells were incubated with 10 μM MG132 (HY-13259, MedChemExpress, Shanghai, China) for 8 h. HCC cells were treated with IP lysis buffer. The total cell lysates were collected and subjected to immunoprecipitation with 2 µg PTEN (ab267787, Abcam) or IgG antibody and 30 µL slurry of Protein G Sepharose (GE Healthcare Life Sciences, Pittsburgh, PA, United States) at 4°C overnight. Western blot analysis was performed with an Anti-Ubiquitin antibody (ab7254, Abcam) to detect the corresponding protein ubiquitination.

The cell counting kit-8 (CCK-8) kit (Beyotime) and Cell-Light™ EdU Apollo^®488 In Vitro Imaging Kit (RIBOBIO, Guangzhou, China) were used for HCC cell proliferation determination as previously described (Shi et al., 2021). For colony formation assay, the logarithmic growth cells (1 × 10³) were seeded into 6-well plates and were grown for a period of 2 weeks. Cell colonies were stabilized with 4% paraformaldehyde, stained with 0.1% crystal violet solution and manually counted.

Glucose consumption and lactate production of HCC cells were detected using the lactate assay kit (ab65331, Abcam) and glucose assay kit (ab136955, Abcam), respectively, according to the manufacturer’s protocols.

The animal studies were approved by the Institutional Animal Care and Use Committee of Xi’an Jiaotong University. Male BALB/c nude mice (4 weeks old) were obtained from the Vital River Laboratory Animal Technology Co. (Beijing, China). 5 × 10⁶ HCCLM3 cells stably transfected with RNF146 shRNA or the negative control were respectively injected into the right armpits of five mice, and subcutaneous tumor size was assessed every week. The mice were sacrificed 4 weeks after injection and the xenograft tumors were removed for IHC with Ki-67 antibody (27309-1-AP, Proteintech) or WB with RNF146 antibody (ab201212, Abcam).

All data in this study were processed by GraphPad Prism 8 (San Diego, CA, United States). Measurement data were shown as mean ± S.D. The Mann–Whitney test and analysis of variance (ANOVA) were applied for the significant test. Survival data were analyzed through the Kaplan-Meier method and log-rank test. p-value less than 0.05 denoted statistically significant.

We analyzed our previous microarray data (GSE155505) of hypoxia-related mRNAs in Hep3B cells. A significant upregulation of RNF146 in Hep3B cells under hypoxic conditions caught our attention (Figure 1A). RT-qPCR and WB analyses consistently confirmed that hypoxia upregulated RNF146 expression in Huh7 and Hep3B cells (p < 0.05, Figures 1B,C). Then, shRNAs were used to downregulate HIF-1α or HIF-2α or HIF1/2α in HCC cells (Figure 1E). The silencing of HIF-1α or HIF-2α or both [double knockdown (DKD)] prominently reduced hypoxia-induced RNF146 expression in HCC cells (p < 0.05, Figures 1D,E). Furthermore, TCGA data analysis using the GEPIA website (http://gepia.cancer-pku.cn/) (Tang et al., 2019) found positive correlations between RNF146 and HIF1/2α in HCC tissues (p < 0.05, Supplementary Figure S1). Our data suggest that hypoxia induces RNF146 expression via HIF-1/2α in HCC cells.

TCGA data analysis using the GEPIA website (http://gepia.cancer-pku.cn/) (Tang et al., 2019) found that RNF146 mRNA expression in HCC was prominently higher than in normal liver tissues (p < 0.0001, Figure 2A). IHC data confirmed the upregulated level of RNF146 protein in HCC tissues compared to adjacent nontumor tissues (p < 0.0001, Figures 2B,C). Moreover, the upregulated levels of RNF146 were also detected in Hep3p, Huh7 and HCCLM3 cells (p < 0.05, Figure 2D). Table 1 showed that HCC patients with large tumors and venous infiltration expressed higher levels of RNF146 (p < 0.05). Survival analysis indicated a close correlation between high RNF146 expression and poor 3-years overall survival of patients with HCC (p = 0.0058, Figure 2E). Thus, these results indicate that RNF146 may be a promising prognostic biomarker for HCC.

Lentivirus-mediated shRNAs markedly downregulated RNF146 in HCCLM3 cells (p < 0.05, Figure 3A). RNF146 knockdown remarkably reduced the viability of HCCLM3 cells (p < 0.05, Figure 3B). The silencing of RNF146 significantly weakened the proliferation of HCCLM3 cells (p < 0.05, Figure 3C). The HCCLM3 cell colonies were obviously decreased by RNF146 silencing (p < 0.05, Figure 3D). Otherwise, RNF146 depletion remarkably decreased the glucose consumption and lactate production of HCCLM3 cells (p < 0.05, Figures 3E,F). RNF146 overexpression prominently enhanced the proliferation and glycolysis of Huh7 cells (p < 0.05, Figure 4). Next, in vivo experiments confirmed that RNF146 knockdown reduced the volume and weight of subcutaneous tumors formed by HCCLM3 cells (p < 0.05, Figures 5A,B). IHC staining of Ki-67 in tumor tissues indicated that the percentage of positive cells in the RNF146 knockdown group was obviously lower than in the control group (p < 0.05, Figure 5C). WB analysis confirmed the downregulated expression of RNF146 in the RNF146 knockdown group compared to the control group (p < 0.05, Figure 5D). Collectively, we consider RNF146 as an oncogene in HCC.

KEGG enrichment pathways analysis suggested that RNF146 might regulate the mTOR signaling pathway in HCC based on TCGA data (Figure 6A and Supplementary Table S1). RNF146 knockdown decreased while RNF146 overexpression increased p-AKT and p-mTOR levels in HCC cells (Figures 6B,C). A previous study has identified RNF146 as a negative regulator of PTEN and activates the PI3K/AKT/mTOR pathway (Li et al., 2015). Interestingly, we found that RNF146 inversely regulated PTEN protein levels in HCC cells (Figures 6B,C). A proteasome inhibitor, MG132, blocked RNF146-induced PTEN downregulation in Huh7 cells (Figure 6D). Moreover, RNF146 silencing obviously reduced the ubiquitination and degradation of PTEN in HCC cells (Figures 6E,F). Therefore, RNF146 promotes PTEN ubiquitination and degradation and activates the AKT/mTOR pathway in HCC.

We intended to verify the involvement of PTEN in RNF146-induced HCC progression. A specific shRNA downregulated PTEN expression in HCCLM3 cells with RNF146 knockdown (Figure 7A). PTEN knockdown recovered RNF146 silencing-induced p-AKT and p-mTOR downregulation in HCCLM3 cells (Figure 7A). We confirmed that PTEN silencing significantly abolished RNF146 knockdown-induced proliferation inhibition of HCCLM3 cells (p < 0.05, Figures 7B,C). Moreover, the effects of RNF146 silencing on HCCLM3 cell colony formation, glucose consumption and lactate production were markedly reversed by PTEN knockdown (p < 0.05, Figures 7D–F). Therefore, PTEN mediates the pro-HCC function of RNF146.