Hepatocellular carcinoma (HCC) is one of the most aggressive malignancies. Elucidating the underlying mechanisms of this disease could provide new therapeutic strategies for treating HCC. Here, we identified a novel role of DEAD-box helicase 24 (DDX24), a member of the DEAD-box protein family, in promoting HCC progression. DDX24 levels were significantly elevated in HCC tissues and were associated with poor prognosis of HCC. Overexpression of DDX24 promoted HCC migration and proliferation in vitro and in vivo, whereas suppression of DDX24 inhibited both functions. Mechanistically, DDX24 bound the mRNA618–624nt of laminin subunit beta 1 (LAMB1) and increased its stability in a manner dependent upon the interaction between nucleolin and the C-terminal region of DDX24. Moreover, regulatory factor X8 (RFX8) was identified as a DDX24 promoter-binding protein that transcriptionally upregulated DDX24 expression. Collectively, these findings demonstrate that the RFX8/DDX24/LAMB1 axis promotes HCC progression, providing potential therapeutic targets for HCC.

Significance:

The identification of a tumor-promoting role of DDX24 and the elucidation of the underlying regulatory mechanism provide potential prognostic indicators and therapeutic approaches to help improve the outcome of patients with hepatocellular carcinoma.

Hepatocellular carcinoma (HCC), the most common subtype of liver malignancy, ranks third in cancer-related deaths worldwide (1, 2). Despite recent therapeutic advances in HCC treatment, recurrence and metastasis are still inevitable and the 5-year survival rate with advanced HCC remains dismal (3). Thus, understanding the molecular mechanisms of HCC tumorigenesis and metastasis are urgently needed in order to precisely target this disease.

The DEAD-box family, characterized by the conserved Asp-Glu-Ala-Asp (DEAD) motif, is the largest group of RNA helicases (4, 5) and is essential in the metabolic processes of RNAs (6–10). Thus, they are identified to be key players in various biological processes such as embryogenesis (11), spermatogenesis (12), cellular growth (13), and tumorigenesis (14). However, studies have suggested important but disparate roles of DEAD-box proteins in cancer. For example, DEAD-box helicase 21 (DDX21), DDX31, and DDX46 are overexpressed and function as oncogenes in various cancer types (15–17). Conversely, DDX3, DDX10, and DDX47 act as tumor suppressors in several cancers including lung and ovarian cancer (18–20), which suggests that the functions of DEAD-box proteins are context-dependent. Previous studies have implicated the function of DDX24 in innate immune signaling (21, 22). In the setting of cancer, DDX24 is potentially a positive regulator of tumor growth, as shown in studies with cancer cell lines (23, 24). However, the expression, function and regulation of DDX24 in HCC are still unclear.

The laminin family members are composed of three nonidentical chains involving α, β, and γ, and each chain can form a short arm respectively (25, 26). Functioning as an activator on cancer cell migration and aggression, laminins have been considered to be adverse prognostic factors in many types of cancer (27–29). As an important member of the laminin family, laminin subunit beta 1 (LAMB1) was shown to be highly expressed in multiple tumors, including HCC, to promote disease progression (30–33). Yet the upstream regulator of this pro-tumor factor remains enigmatic.

In this study, we revealed that DDX24 promoted HCC by binding to LAMB1 mRNA and increasing its stability in a nucleolin (NCL)-dependent manner. We then identified regulatory factor X8 (RFX8) as a novel transcriptional factor that upregulated the expression of DDX24 and LAMB1. Overall, our study demonstrated the regulatory mechanisms underlying DDX24-driving HCC progression.

Cell lines

Hep3B, PLC/PRF/5, SNU398, and SNU449 were obtained from the ATCC. Huh 7 was obtained from the Japanese Cancer Research Resources Bank (JCRB). L02 was obtained from the Chinese Academy of Sciences Cell Bank. Cells were authenticated by short tandem repeat (STR) profiling (Supplementary data) and were confirmed to be Mycoplasma free before use.

qRT-PCR

Total RNA from the cell lines used in this study was extracted with E.Z.N.A. Real-time PCR was performed to analyze gene expression using All-in-One qPCR Mix (GeneCopoeia), followed by detection via Bio-Rad CFX96, and analysis using Bio-Rad Manager software (Bio-Rad). Expression levels were normalized to that of the housekeeping gene GAPDH, and each sample was assessed in triplicate. The related primers were listed in Supplementary Table S1.

Western blotting

Whole-cell lysates (WCL) were prepared using RIPA lysis buffer and protease/phosphatase Inhibitor Cocktail. Nuclear and cytoplasmic fractions were extracted and isolated using a Nuclear/Cytoplasmic Isolation kit (BestBio). Protein concentration was determined using a bicinchoninic acid (BCA) assay (Beyotime Biotechnology). Proteins were separated via SDS-PAGE gel and transferred onto polyvinylidene difluoride (PVDF) membranes for detection. Antibody information is showed in Supplementary Table S2.

Clinical samples

Fresh tumor and nontumor liver tissues were collected from 11 patients (Supplementary Table S3) with HCC who underwent hepatectomy at the Fifth Affiliated Hospital of Sun Yat-Sen University (Zhuhai, China). The resected samples used in IHC were formalin fixed and paraffin embedded and identified by two pathologists independently. The samples of cohorts 1 to 3 were collected from 2007 to 2009, 2006 to 2007, 2007 to 2008 respectively. The clinical characteristics of all patients in three cohorts are listed in Supplementary Table S4. This study was approved by the approval of the Medical Ethics Committee of The Fifth Affiliated Hospital of Sun Yat-sen University (2018-K151–1), and all patients provided written informed consent.

Cell viability assay

Cellular viability was assayed using Cell Counting Kit-8 (CCK-8; KeyGEN BioTECH). HCC cells were seeded at a density of 3 × 103 cells in 96-well plates, using 100-μL cell suspension per well. After 72 hours of culture, absorbance was measured at 490 nm per well using a microplate reader.

Cell migration assay

Cell migration assays were performed with Transwell filter chambers (Becton Dickinson). 3 × 104 HCC cells in 200 μL serum-free DMEM were seeded in the upper chamber of a transwell and 400 μL medium supplemented with 20% FBS was added to the lower chamber. Cells migrated through the membrane were fixed and stained. Three random fields were counted under the light microscope.

Generation of lentiviral particles and cell transduction

DDX24 knockdown or overexpression in cell lines was established by infecting cells with a lentiviral vector purchased from GeneCopoeia. HCC cells were infected with recombinant lentivirus transducing units with 0.8 μg/mL polybrene. Clones were selected after 2 weeks using 1 μg/mL puromycin. Efficiency of RNAi was assessed via qRT-PCR and Western blotting.

Immunofluorescence staining

HCC cells were seeded on chamber slides and cultured for 24 hours. Cells were washed with PBS, fixed with precooled paraformaldehyde for 10 minutes at 4°C, permeabilized with 1% TritonX-100 (Solarbio Life Science) in PBS for 15 minutes, incubated in blocking solution for 1 hour at room temperature, and incubated with specific primary antibodies overnight at 4°C. Then, the samples were washed three times with PBS, incubated with secondary antibodies for 1 hour at room temperature in dark place, washed three times with PBS, and incubated for 10 minutes in the dark with Fluoroshield Mounting Medium with DAPI (Abcam). Images were acquired using a fluorescence inverted electron microscope (IX73) or confocal microscope.

IHC staining

Samples of HCC tissues were acquired during surgery at the Fifth Affiliated Hospital of Sun Yat-sen University. IHC was carried out using streptavidin–peroxidase-conjugated method. Briefly, each tissue section was deparaffinized, rehydrated, immersed in antigen retrieval solution, boiled at 100°C for 10 minutes, and incubated with a peroxidase inhibitor. Then, nonspecific binding was blocked with normal goat serum, and tissue sections were incubated overnight at 4°C with the primary antibodies indicated.

mRNA stability assay

Cells were pretreated with DDX24 short hairpin RNA (shRNA) or negative control shRNA before the addition of actinomycin-D (used at 100 μg/mL final concentration; TargetMol). Total mRNA was extracted at 0 to 10 hours and measured by real-time RT-PCR. The values presented are relative to those obtained using control cells at the time of actinomycin D addition.

Streptavidin-agarose pulldown assay

The DDX24 promoter binding proteins were identified by streptavidin-agarose pulldown assay. Briefly, 200 μg nuclear protein was extracted from different HCC cell lines and incubated at 4°C overnight with 2 μg biotin-labeled double-stranded DNA probes corresponding to nucleotides −745 to +61 of the DDX24 promoter region and 20 μL streptavidin-agarose beads (Sigma-Aldrich). The mixture was centrifuged to pulldown the DNA-protein complex, after which, an SDS-PAGE gel was used to separate the proteins, and silver staining was used to detect the proteins after electrophoretic separation. Candidate protein bands were then trypsinized and digested with the MS-grade trypsin solution (Promega). Finally, the digested peptides from the candidate proteins were analyzed by MALDI-TOF/TOF mass spectrometry.

Dual-luciferase reporter assay

HCC cells in 96-well plates (5 × 103 cells/well) were cotransfected with RFX8 or negative control, and 20 ng firefly luciferase reporter plasmid using Lipofectamine 3,000 (Invitrogen) per manufacturer's instructions. At 72 hours after transfection, cell lysates were analyzed using the dual-luciferase reporter assay system (Promega), according to the manufacturer's instructions.

Chromatin immunoprecipitation assay

HCC cells were fixed with 1% formaldehyde, and then the cross-linking reaction was quenched by adding 100 μL 1.375 mol/L glycine per milliliter culture media. Samples of the cells were sonicated on ice to shear the DNA. Cell lysates were immunoprecipitated with anti-RFX8 antibody (Genetex) or nonimmune rabbit IgG (Cell Signaling Technology). DNA fragments were purified using spin columns (Qiagen), and PCR was used to evaluate the enrichment of the DDX24 promoter region using the following primer pair -Forward: 5′-TGGCCCCTCCCTCGGGTTAC-3′, Reverse: 5′-TGAAGGGGCAGGACG GGTGC-3′. PCR products were resolved by electrophoresis in a 2% agarose gel and visualized by gel-red staining.

RNA immunoprecipitation sequencing

HCC cells were harvested using a cell scraper. Then, the cells were pelleted, resuspended in ice-cold polysomal lysis buffer [10 mmol/L HEPES, pH 7.0; 100 mmol/L KCl; 5 mmol/L MgCl2; 0.5% NP40; 1 mmol/L DTT; 100 U/mL RNase Inhibitor (Takara); 1×Protease Inhibitor Cocktail (Roche); and 0.4 mmol/L RVC (New England Biolabs)], and incubated on ice for 15 minutes. After the lysates were centrifuged at 15,000 g for 15 minutes, the supernatants were precleared with Dynabeads Protein G (Invitrogen). Each lysate was then diluted in NT2 buffer [50 mmol/L Tris, pH 7.4; 150 mmol/L NaCl; 1 mmol/L MgCl2; 0.05% NP40; 1 mmol/L DTT; 100 U/mL RNase Inhibitor (Takara); 1×Protease Inhibitor Cocktail (Roche); and 20 mmol/L EDTA]. One percent of the supernatant was saved as input and the rest of the lysate was utilized for RIP assay at 4°C overnight using 5 μg DDX24 antibody (15769–1-AP; Proteintech) or a corresponding control IgG (NI01–100UG; Rabbit IgG; EMD Millipore). The next day, Dynabeads Protein G was added, and the mixtures were incubated at 4°C for 3 hours, followed by washes with NT2 buffer. A quarter of the RIP material was used for Western blotting, and the rest was utilized for RNA extraction. Fold enrichment of RNAs was detected by next generation sequencing or qPCR. Genes with base mean ≥500, IP/input ≥2, P < 0.05 were identified as target genes.

RNA pulldown

The RNA pulldown kit (BersinBio) was used for the assay according to the manufacturer's protocol. In brief, about 2 × 107 cells were harvested and lysed on ice, then whole-protein lysate was collected. Streptavidin magnetic beads were prewashed by TES (Tris-HCl, EDTA, SDS) buffer. Biotinylated RNA probes were bound to the beads at 25°C for 30 minutes. The RNA-bound beads were separated on magnetic stand and washed with TES buffer twice. The lysate was incubated with RNA-bound beads on a rotator at 25°C for 2 hours. The beats were washed briefly four times at 4°C. The beads were separated on a magnetic stand, and the supernatant was then discarded. RNA-binding protein complexes were eluted using protein elution buffer and DTT at 37°C for 2 hours. The retrieved protein was analyzed by Western blot.

Animal studies

All animal procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH publications nos. 80–23, revised 1996) and Institutional Ethical Guidelines for Animal Experiments developed by the Sun Yat-sen University. Animals (Vital River Laboratory Animal Technology Co., Ltd.) were maintained under pathogen-free conditions in Guangdong Provincial Key Laboratory of Biomedical Imaging. BALB/c female nude mice aged 4 to 6 weeks were subcutaneously injected with 1 × 106 indicated cells. The length (L), width (W), and height (H) of each tumor were measured with calipers, and tumor volume (V) was calculated as follows: V = πLWH/6. At the end of the experiment, the mice were humanely sacrificed in the morning, tumors were excised, imaged, and measured. For the lung metastatic model, BALB/c female nude mice were each given intravenous tail vein injection with 1 × 106 indicated HCC cells. All mice were sacrificed in the morning after 8 weeks, and the lungs were harvested. Tumor nodules on the lung surfaces were counted, excised, and embedded in paraffin.

Statistics

Results are shown as mean ± SD of at least three independent experiments. Statistical analysis was performed using GraphPad Prism software (version 7.0). Differences between groups were evaluated using Student t test unless otherwise specified (paired t test or χ2 test). Cumulative survival was evaluated using the Kaplan–Meier method (log–rank test). Statistical significance is indicated as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001; not significant (NS).

Data availability statement

The data generated in this study are available within the article and its supplementary data files. Raw sequencing data from high-throughput sequencing are available in a public, open access repository. The links are provided below:

RNA sequencing (RNA-seq), https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE145635; RNA immunoprecipitation sequencing (RIP-seq), https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSE145632.

The mass spectrometry proteomics data are available at: RFX8, http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD022650; NCL, http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD023321.

Ethics approval and consent to participate

This work was done with the approval of the Medical Ethics Committee of The Fifth Affiliated Hospital of Sun Yat-sen University (2018-K151–1). The mice used in this study were cared for and handled according to animal protocols approved by the Animal Care Committee of The Fifth Affiliated Hospital of Sun Yat-sen University (00026).

DDX24 is upregulated in HCC tissues and associated with poor prognosis in patients with HCC

To survey the expression profile of DEAD-box genes in HCC, we first resorted to The Cancer Genome Atlas (TCGA) database. In the 424 tissue samples (ENSG00000089737), including 374 from patients with HCC and 50 from noncancerous liver tissues, we identified a list of differentially expressed DDX genes (Fig. 1A). Among them, DDX24 was significantly elevated in HCC tissues with its function and regulation largely unexplored (Fig. 1B). Furthermore, DDX24 was upregulated in patients with HCC with stage III/IV (Fig. 1C). These analyses pointed to DDX24 as a candidate gene to be studied. We next validated the expression of DDX24 protein in 11 paired HCC tissues that were acquired from the Fifth Affiliated Hospital of Sun Yat-sen University, which confirmed the TCGA results (Fig. 1D). Then we examined five HCC cell lines together with a nonmalignant hepatic cell line L02 and demonstrated that DDX24 expression was higher in all the HCC cell lines tested than L02 (Fig. 1E).

To explore the potential role of DDX24 in HCC, we analyzed its expression in tissues from three individual cohorts [cohort 1 (n = 92), cohort 2 (n = 88), and cohort 3 (n = 90)] of patients with HCC. The expression of DDX24 was significantly increased in the tumor tissues compared with that in the nontumor tissues (Fig. 1F). We further analyzed the correlation between DDX24 expression and clinicopathologic features in the three cohorts of patients (Supplementary Table S4). Results showed that DDX24 expression was strongly correlated with tumor size (P = 0.0249) and T stage (P = 0.048; Supplementary Table S5). Also, higher DDX24 level was more subject to tumor recurrence (cohort 2, P < 0.0001; cohort 3, P < 0.0001) and a greater number of cirrhosis nodules (cohort 3, P = 0.0297; Supplementary Tables S6–S8). Kaplan–Meier survival analysis revealed that patients with high levels of DDX24 had significantly lower overall survival (OS) and disease-free survival (DFS) than those with low DDX24 expression (Fig. 1G and H; Supplementary Fig. S1A–S1E). Multivariate analysis indicated that DDX24 expression was an independent prognostic factor for the OS of patients with HCC [95% confidence interval (CI), 2.724–5.596; P < 0.001; Table 1]. Taken together, these findings suggest that upregulation of DDX24 is strongly associated with HCC progression.

DDX24 promotes HCC metastasis and growth

To investigate the function of DDX24 in HCC, we performed RNA-seq in Hep3B with or without knocking down DDX24 expression. The results of RNA-seq (GSE145635) suggested that the cellular processes associated with DDX24 involved cell motility and growth (Fig. 2A). We then used Hep3B and SNU398, two cell lines with metastatic properties, to study the function of DDX24 in HCC. Indeed, knockdown of DDX24 inhibited HCC cell migration (Fig. 2B and C) and reduced the number of metastatic nodules in the murine model for lung metastasis (Fig. 2D). Reversely, overexpression of DDX24 promoted HCC cell migration (Fig. 2E and F) and increased the number of lung metastatic nodules (Fig. 2G). Furthermore, as suggested by the RNA-seq results, we confirmed that downregulation of DDX24 suppressed cell proliferation (Fig. 2H) and tumor growth (Fig. 2IK), whereas upregulation of DDX24 elevated both (Fig. 2LO). These results suggest that DDX24 promotes HCC metastasis and growth.

DDX24 binds to and stabilizes LAMB1 mRNA

Because DDX24 is an RNA-binding protein, we performed RIP-seq (GSE145632) to probe the downstream targets of DDX24 (Fig. 3A). Protein-coding mRNAs predominated in the DDX24-bound RNAs (Supplementary Fig. S2A). The top 20 enriched pathways indicated by Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis included the regulation of cell morphogenesis and cell growth pathways (Supplementary Fig. S2B). Combining RNA-seq and RIP-seq data, seven candidate genes stood out for further investigation (Fig. 3B and C). Among these genes, the mRNA of LAMB1 was most significantly downregulated upon DDX24 knockdown (Fig. 3D). On the other hand, overexpression of DDX24 resulted in an elevated mRNA level of LAMB1 (Fig. 3E). Moreover, DDX24 positively regulated LAMB1 at the protein level (Fig. 3F), verifying that LAMB1 mRNA was the direct target of DDX24 protein.

We then sought to identify the specific binding region of DDX24 on the LAMB1 mRNA. We adopted Homer motif analysis to DDX24 enhanced cross-linking and immunoprecipitation (eCLIP)-seq data (ENCODE accession ID ENCFF142OCF; ref. 34). The most preferential binding motif in mRNA transcripts of DDX24 was shown in Fig. 3G. Next, we used RSAT website tool (http://rsat.sb-roscoff.fr/matrix-scan.cgi) to scan LAMB1 mRNA sequence with the shown motif by position-specific weight matrix (PSWM). This analysis identified five putative binding sites in the LAMB1 mRNA (Fig. 3G). Subsequently, we constructed five corresponding probes for RNA pulldown assay (Fig. 3H) and showed that the region 138 to 1029nt (including site 1) in LAMB1 mRNA was critical for DDX24 binding (Fig. 3I). To further validate the binding region, we generated a mutant probe by depleting site 1. As expected, the binding of LAMB1 mRNA with DDX24 was completely abolished when the region of 618 to 624nt was missing (Fig. 3J). Moreover, the stability of LAMB1 mRNA following depletion of region 1 (618–624nt) was reduced (Fig. 3K), indicating that DDX24 bound to and regulated LAMB1 mRNA by increasing its stability. To test this hypothesis, we performed RNA stability assay and found that reducing the level of DDX24 resulted in the faster degradation of LAMB1 mRNA (Fig. 3L, Supplementary Table S9). Therefore, DDX24 regulates LAMB1 by stabilizing LAMB1 mRNA.

To verify LAMB1 expression in HCC patient samples, we showed that LAMB1 expression was significantly increased in HCC tissues compared with nontumor tissues at the protein level (Fig. 3M). Analysis of TCGA database (n = 424, ENSG00000091136) also revealed that LAMB1 was upregulated in HCC tissues at the mRNA level (Fig. 3N). Moreover, the expression of LAMB1 protein was positively correlated with that of DDX24 (Fig. 3O). At the mRNA level, similar results were observed in the TCGA cohort (Fig. 3P). Furthermore, the OS of patients with HCC who had low expression of LAMB1 or DDX24 or both was much longer than those whose LAMB1 and DDX24 were highly expressed (Fig. 3Q; Supplementary Table S10). These findings demonstrate that DDX24 binds to LAMB1 mRNA and upregulates its expression in HCC.

DDX24 promotes HCC progression via LAMB1 signaling

To test whether DDX24 promotes HCC progression via LAMB1 signaling, we first examined whether overexpression of LAMB1 could rescue the inhibitory effects caused by DDX24 knockdown in HCC cells. Our results showed that the inhibited migration and viability of Hep3B and SNU398 cells upon DDX24 knockdown were largely reverted by overexpression of LAMB1 (Fig. 4AC). Furthermore, knocking down LAMB1 abrogated the promotive effect on HCC cell migration (Fig. 4DF) and tumor growth (Fig. 4GJ) mediated by DDX24 overexpression in vitro and in vivo. In addition, the phosphorylation of mTOR, AKT, and ERK were downregulated when LAMB1 was knocked down (Fig. 4K). Together, these findings illustrate that DDX24 regulates HCC metastasis and growth via LAMB1-mediated AKT/mTOR/ERK signaling.

DDX24 stabilizes LAMB1 mRNA by interacting with NCL

To identify proteins that interact with DDX24 in HCC, we performed immunoprecipitation combined with mass spectrometry in DDX24-Flag overexpressing Hep3B cells. NCL, a nucleolus protein, with the peptide sequence of GFGFVDFNSEEDAK, was detected (Fig. 5A; Supplementary Fig. S3). The interaction between DDX24 and NCL was further confirmed by coimmunoprecipitation assay in Hep3B cells (Fig. 5B). Moreover, DDX24-mediated LAMB1 upregulation was compromised when NCL was abolished, suggesting that DDX24-regulated LAMB1 expression is NCL-dependent (Fig. 5C). Because the interaction between DDX24 and NCL has not been previously reported, we extended our analysis to map the NCL-interacting domain on DDX24 protein by generating a series of Flag-tagged DDX24 truncation mutants as shown in Fig. 5D. We found that the 159 amino acids of the C-terminal sequence of DDX24 mediated its interaction with NCL (Fig. 5E and F). Functionally, silencing NCL abrogated the cellular effects induced by DDX24 overexpression (Fig. 5G and H). These results demonstrate that DDX24 stabilizes LAMB1 dependent of its interaction with NCL.

RFX8 transcriptionally upregulates DDX24 in HCC

To elucidate the upstream regulatory mechanism of DDX24 in HCC, a 5′-biotin–labeled 806-bp DNA probe targeting the −745 to +61 region of DDX24 promoter (Fig. 6A), was synthesized to screen and identify DDX24 promoter-binding proteins (Fig. 6B). Data acquired using two individual cell lines (Hep3B and SNU449) displayed the same single protein band between 55kDa and 70 kDa that was markedly enhanced in HCC cells but not in immortalized hepatic cell line L02 (Fig. 6C). The results of matrix-assisted laser desorption ionization-time of flight (MALDI-TOF/TOF) mass spectrometry indicated that this protein possessed the specific sequence of DILRNVR (Fig. 6D; Supplementary Fig. S4A; Supplementary Table S11). According to the proteomics database (https://blast.ncbi.nlm.nih.gov/Blast.cgi), this sequence belonged to the DNA binding protein RFX8 (64 kDa). Immunofluorescence analysis further confirmed that RFX8 localized to the cell nucleus (Fig. 6E).

To verify whether RFX8 can bind to DDX24 promoter, we applied the nuclear protein/DNA complex pulldown assay with a 5′-biotin labeled DDX24 promoter probe (DPP) or nonspecific probe (NSP). As shown in Fig. 6F, RFX8 was detected by the DPP, but not by the NSP. Furthermore, in order to determine whether RFX8 directly regulated the transcription of DDX24 gene, we generated a reporter construct containing the core promoter region of DDX24 (−745 to +71) located upstream of the dual-luciferase gene and showed that knockdown of RFX8 significantly reduced the activity of DDX24 promoter (Fig. 6G and H; Supplementary Fig. S4B). Overexpression of RFX8 dramatically enhanced DDX24 promoter activity (Fig. 6I; Supplementary Fig. S4C).

In order to pinpoint the specific binding region on DDX24 promoter by RFX8, a series of reporter constructs containing deletions of the DDX24 5′-flanking regions were cloned upstream of the dual-luciferase gene (Fig. 6H). Our results showed that RFX8 failed to enhance the luciferase activity of the reporter without the region of −335 to −135 (Fig. 6I; Supplementary Fig. S4C). Thus, location of the principal DDX24 transcriptional control element was narrowed down to a region spanning the nucleotides −335 to −135. Further sequence analysis revealed one putative regulatory factor binding to the X-box (RFX) transcription factors binding site (5′-CGTTACTGT-3′) in this region (Fig. 6J). Dual-luciferase reporter assay proved that RFX8 could not enhance the luciferase activity of the reporter when the putative binding site was mutated (5′-CGTTACTGT-3′ to 5′-ATGACAGCA-3′; Fig. 6K and L). To further validate the interaction between RFX8 protein and DDX24 promoter, chromatin immunoprecipitation (ChIP) was performed in three HCC cell lines. As shown in Fig. 6M, RFX8 protein bound to the endogenous DDX24 promoter in all the cell lines tested. These data elucidate that RFX8 binds to DDX24 promoter and activates DDX24 transcription.

To validate the role of RFX8 in transcription regulation, we demonstrated that knockdown of RFX8 suppressed the expression of DDX24 and LAMB1, whereas overexpressing RFX8 promoted both mRNA and protein levels of DDX24 and LAMB1 (Fig. 6NP), suggesting that RFX8 worked upstream of the DDX24/LAMB1 axis. To confirm the clinical significance of RFX8 in HCC, we found that RFX8 expression was significantly enhanced in HCC tissues compared with that in nontumor tissues (Supplementary Fig. S4D). Moreover, the expression of RFX8 was positively correlated with the expression of DDX24 at the protein level (Supplementary Fig. S4E). Similar results were observed at the mRNA level in the TCGA cohort (Supplementary Fig. S4F). Here we demonstrate that RFX8 is a DDX24 promoter-binding protein that transcriptionally upregulates DDX24 expression in HCC.

In this study, we identified DDX24 as a novel HCC promoter. Elevated expression of DDX24 in HCC tissues correlated with poor prognosis. Suppressing DDX24 expression inhibited HCC growth and metastasis in vitro and in vivo. We also revealed that DDX24 protein bound to and stabilized the mRNA of LAMB1 in an NCL-dependent manner. We further identified RFX8 as the transcription factor of DDX24, which enhanced the transcriptional activity of DDX24, leading to the upregulated DDX24 expression in HCC. Together, our results indicate that DDX24 acts as a protumor factor in HCC and might be a prognostic indicator and therapeutic target for HCC.

Recent studies have suggested disparate roles of DEAD-box proteins in HCC. For example, DDX3 suppressed HCC by modulating tumor-suppressive miRNAs and transcriptional regulation activity of p21waf1/cip1 (18, 35, 36); in contrast, DDX17 and DDX39 promotes HCC progression by inhibiting Klf4 transcriptional activity and activating Wnt/β-catenin pathway respectively (37, 38). Only a handful of literature is available regarding the function of DDX24, which relates to innate immune signaling and tumorigenesis in vitro (21–24, 39). Our results agree with previous studies in that DDX24 functions as a positive regulator of tumor growth, with a novel and comprehensive demonstration of the mechanism involved.

NCL, accounting for approximately 10% of the protein content in cell nucleolus, participates in RNA metabolism and gene transcription and plays significant roles in many physiological processes such as modulating the proliferation and migration of cancer cells (40). It has been reported that DEAD-box proteins often function in the form of a complex with nucleolus proteins (31). For instance, Fukawa and colleagues found that DDX31 interacted with nucleophosmin (NPM1) in the nucleoli and regulated the p53-HDM2 pathway and rRNA genes in renal cell carcinoma (15). Another study revealed that DDX31 may play a crucial role in EGFR stabilization via its binding to NCL in muscle-invasive bladder cancer (41). In this article, we found that the C-terminal sequence of DDX24 mediated its interaction with NCL in HCC. As a multifunctional protein, NCL has been demonstrated to play an oncogenic role in HCC (42). However, the mechanism by which NCL regulates HCC remained unclear. Here, we identified that NCL interacted with DDX24 to regulate downstream LAMB1, indicating a working model by which NCL contributes to protumor activities.

LAMB1, a DDX24 downstream factor proved in this study, belongs to a large family of extracellular matrix glycoproteins (43). The biological functions of laminin-family members are diverse, including cell growth, adhesion, differentiation, and resistance to apoptosis (44). The role of LAMB1 in cancer has been revealed previously (32, 45). For example, a recent study has shown that HCC tumor progression is activated by PDGFRα-La/SSB-LAMB1 axis (32). In this article, we found an alternative way to activate LAMB1 by which DDX24 protein directly bound to and stabilized the RNA of LAMB1, resulting in HCC promotion.

RFX8 is the least studied member in the RFX family (46), despite that the RFX family proteins regulate numerous downstream genes that are essential in cellular and developmental processes (47). The influence of RFX family proteins on HCC had rarely been examined. Here, for the first time, we identified RFX8 as a transcriptional regulator of DDX24 in HCC and that knockdown of RFX8 suppressed DDX24 expression, which inhibited HCC development.

In summary, we identified DDX24 as a novel prognostic biomarker for HCC and uncovered its regulatory mechanisms. Our findings provide a series of novel targets that can be exploited therapeutically against HCC.

T. Liu reports grants from the National Natural Science Foundation of China and grants from the Guangdong Basic and Applied Basic Research Foundation during the conduct of the study; in addition, T. Liu has a patent for ZL202010839780.4 licensed to H. Shan, H. He, T. Liu, H. Gan, and S. He. H. Gan reports a patent for ZL202010839780.4 licensed to H. Shan, H. He, T. Liu, H. Gan, and S. He. S. He reports a patent for ZL202010839780.4 licensed to H. Shan, H. He, T. Liu, H. Gan, and S. He. H. He reports grants from National Natural Science Foundation of China and grants from Guangdong Province during the conduct of the study; in addition, H. He has a patent for ZL202010839780.4 licensed to H. Shan, H. He, T. Liu, H. Gan, and S. He. H. Shan reports a patent for ZL202010839780.4 licensed to H. Shan, H. He, T. Liu, H. Gan, and S. He. No disclosures were reported by the other authors.

H. Shan: Conceptualization, supervision, funding acquisition, project administration. T. Liu: Conceptualization, data curation, funding acquisition, methodology, writing–original draft. H. Gan: Data curation, methodology, writing–original draft. S. He: Formal analysis, methodology. J. Deng: Data curation, methodology. X. Hu: Data curation, methodology. L. Li: Data curation, methodology. L. Cai: Data curation, software, methodology. J. He: Data curation. H. Long: Data curation, methodology. J. Cai: Software, methodology. H. Li: Data curation, methodology. Q. Zhang: Data curation, methodology. L. Wang: Data curation, writing–original draft. F. Chen: Data curation. Y. Chen: Data curation. H. Zhang: Data curation, methodology. J. Li: Conceptualization, resources, supervision. L. Yang: Resources, data curation. Y. Liu: Methodology. J. Yang: Software. D. Kuang: Conceptualization, resources, supervision, project administration, writing–review and editing. P. Pang: Conceptualization, project administration, writing–review and editing. H. He: Conceptualization, supervision, project administration, writing–review and editing.

The authors acknowledge grants from the National Natural Science Foundation of China (grant nos. 81620108017, 81872113, 81802936, 82025016), the Guangdong Basic and Applied Basic Research Foundation (grant nos. 2021A1515010974, 2020A1515010016), the Natural Science Foundation of Guangdong Province (grant nos. 2017A030313873), Collaborative Project from Guangdong Province (grant no. 2020A0505100028), the Department of Science and Technology of Guangdong Province (grant nos. 2018B030322006), and the 5010 Project of Clinical Medicine Research, Sun Yat-sen University (grant no. 2018011).

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.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

1.
Siegel
RL
,
Miller
KD
,
Jemal
A
.
Cancer statistics, 2020
.
CA Cancer J Clin
2020
;
70
;
7
30
.
2.
Lin
X
,
Li
AM
,
Li
YH
,
Luo
RC
,
Zou
YJ
,
Liu
YY
, et al
.
Silencing MYH9 blocks HBx-induced GSK3β ubiquitination and degradation to inhibit tumor stemness in hepatocellular carcinoma
.
Signal Transduct Target Ther
2020
;
5
;
13
.
3.
Ma
XL
,
Hu
B
,
Tang
WG
,
Xie
SH
,
Ren
N
,
Guo
L
, et al
.
CD73 sustained cancer-stem-cell traits by promoting SOX9 expression and stability in hepatocellular carcinoma
.
J Hematol Oncol
2020
;
13
;
11
.
4.
Linder
P
,
Jankowsky
E
,
From unwinding to clamping - the DEAD box RNA helicase family
.
Nat Rev Mol Cell Biol
2011
;
12
;
505
16
.
5.
Linder
P
,
Fuller-Pace
FV
,
Looking back on the birth of DEAD-box RNA helicases
,
Biochim Biophys Acta
2013
;
1829
;
750
5
.
6.
Jarmoskaite
I
,
Russell
R
,
DEAD-box proteins as RNA helicases and chaperones.
Wiley Interdiscip Rev RNA
2011
;
2
;
135
52
.
7.
Tauber
D
,
Tauber
G
,
Khong
A
,
Van Treeck
B
,
Pelletier
J
,
Parker
R
,
Modulation of RNA condensation by the DEAD-box protein eIF4A
,
Cell
2020
;
180
;
411
26
.
8.
Popow
J
,
Jurkin
J
,
Schleiffer
A
,
Martinez
J
,
Analysis of orthologous groups reveals archease and DDX1 as tRNA splicing factors
,
Nature
2014
;
511
;
104
7
.
9.
Xing
YH
,
Yao
RW
,
Zhang
Y
,
Guo
CJ
,
Jiang
S
,
Xu
G
, et al
.
SLERT regulates DDX21 rings associated with pol I transcription
.
Cell
2017
;
169
;
664
78
.
10.
Calo
E
,
Flynn
RA
,
Martin
L
,
Spitale
RC
,
Chang
HY
,
Wysocka
J
.
RNA helicase DDX21 coordinates transcription and ribosomal RNA processing
.
Nature
2015
;
518
;
249
53
.
11.
Konishi
T
,
Uodome
N
,
Sugimoto
A
.
The Caenorhabditis elegans DDX-23, a homolog of yeast splicing factor PRP28, is required for the sperm-oocyte switch and differentiation of various cell types
.
Dev Dyn
2008
;
237
;
2367
77
.
12.
Lee
R
,
Lee
WY
,
Park
HJ
,
Ha
WT
,
Woo
JS
,
Chung
HJ
, et al
.
Stage-specific expression of DDX4 and c-kit at different developmental stages of the porcine testis
.
Anim Reprod Sci
2018
;
190
;
18
26
.
13.
Cruciat
CM
,
Dolde
C
,
de Groot
RE
,
Ohkawara
B
,
Reinhard
C
,
Korswagen
HC
, et al
.
RNA helicase DDX3 is a regulatory subunit of casein kinase 1 in Wnt-β-catenin signaling
.
Science
2013
;
339
;
1436
41
.
14.
Robinson
G
,
Parker
M
,
Kranenburg
TA
,
Lu
C
,
Chen
X
,
Ding
L
, et al
.
Novel mutations target distinct subgroups of medulloblastoma
.
Nature
2012
;
488
;
43
48
.
15.
Fukawa
T
,
Ono
M
,
Matsuo
T
,
Uehara
H
,
Miki
T
,
Nakamura
Y
, et al
.
DDX31 regulates the p53-HDM2 pathway and rRNA gene transcription through its interaction with NPM1 in renal cell carcinomas
.
Cancer Res
2012
;
72
;
5867
77
.
16.
Zhang
Y
,
Baysac
KC
,
Yee
LF
,
Saporita
AJ
,
Weber
JD
.
Elevated DDX21 regulates c-Jun activity and rRNA processing in human breast cancers
.
Breast Cancer Res
2014
;
16
;
449
.
17.
Jiang
F
,
Zhang
D
,
Li
G
,
Wang
X
.
Knockdown of DDX46 inhibits the invasion and tumorigenesis in osteosarcoma cells
.
Oncol Res
2017
;
25
;
417
25
.
18.
Correction
:
reduced p21WAF1/CIP1 via alteration of p53-DDX3 pathway is associated with poor relapse-free survival in early-stage human papillomavirus-associated lung cancer
.
Clin Cancer Res
2015
;
21
;
935
.
19.
Gai
M
,
Bo
Q
,
Qi
L
.
Epigenetic down-regulated DDX10 promotes cell proliferation through Akt/NF-κB pathway in ovarian cancer
.
Biochem Biophys Res Commun
2016
;
469
;
1000
5
.
20.
Lee
JH
,
Rho
SB
,
Chun
T
.
GABAA receptor-associated protein (GABARAP) induces apoptosis by interacting with DEAD (Asp-Glu-Ala-Asp/His) box polypeptide 47 (DDX 47)
.
Biotechnol Lett
2005
;
27
;
623
8
.
21.
Ma
Z
,
Moore
R
,
Xu
X
,
Barber
GN
.
DDX24 negatively regulates cytosolic RNA-mediated innate immune signaling
.
PLoS Pathog
2013
;
9
;
e1003721
.
22.
Ma
J
,
Rong
L
,
Zhou
Y
,
Roy
BB
,
Lu
J
,
Abrahamyan
L
, et al
.
The requirement of the DEAD-box protein DDX24 for the packaging of human immunodeficiency virus type 1 RNA
,
Virology
2008
;
375
;
253
64
.
23.
Oliver
D
,
Ji
H
,
Liu
P
,
Gasparian
A
,
Gardiner
E
,
Lee
S
, et al
.
Identification of novel cancer therapeutic targets using a designed and pooled shRNA library screen
.
Sci Rep
2017
;
7
;
43023
.
24.
Shi
D
,
Dai
C
,
Qin
J
,
Gu
W
.
Negative regulation of the p300-p53 interplay by DDX24
.
Oncogene
2016
;
35
;
528
36
.
25.
Fahey
B
,
Degnan
BM
.
Origin and evolution of laminin gene family diversity
.
Mol Biol Evol
2012
;
29
;
1823
36
.
26.
Schéele
S
,
Nyström
A
,
Durbeej
M
,
Talts
JF
,
Ekblom
M
,
Ekblom
P
.
Laminin isoforms in development and disease
.
J Mol Med
2007
;
85
;
825
36
.
27.
Rousselle
P
,
Scoazec
JY
.
Laminin 332 in cancer: when the extracellular matrix turns signals from cell anchorage to cell movement
.
Semin Cancer Biol
2020
;
62
;
149
65
.
28.
Yamamoto
H
,
Kitadai
Y
,
Yamamoto
H
,
Oue
N
,
Ohdan
H
,
Yasui
W
, et al
.
Laminin gamma2 mediates Wnt5a-induced invasion of gastric cancer cells
.
Gastroenterology
2009
;
137
;
242
52
.
29.
Patarroyo
M
,
Tryggvason
K
,
Virtanen
I
.
Laminin isoforms in tumor invasion, angiogenesis and metastasis
.
Semin Cancer Biol
2002
;
12
;
197
207
.
30.
Simon-Assmann
P
.
The laminin family: founding members of the basement membrane
.
Cell Adh Migr
2013
;
7
;
44
47
.
31.
Chen
Q
,
Lu
G
,
Cai
Y
,
Li
Y
,
Xu
R
,
Ke
Y
, et al
.
MiR-124–5p inhibits the growth of high-grade gliomas through posttranscriptional regulation of LAMB1
.
Neuro-oncol
2014
;
16
;
637
51
.
32.
Govaere
O
,
Petz
M
,
Wouters
J
,
Vandewynckel
YP
,
Scott
EJ
,
Topal
B
, et al
.
The PDGFRα-laminin B1-keratin 19 cascade drives tumor progression at the invasive front of human hepatocellular carcinoma
.
Oncogene
2017
;
36
;
6605
16
.
33.
Govaere
O
,
Komuta
M
,
Berkers
J
,
Spee
B
,
Janssen
C
,
de Luca
F
, et al
.
Keratin 19: a key role player in the invasion of human hepatocellular carcinomas
.
Gut
2014
;
63
;
674
85
.
34.
Van Nostrand
EL
,
Freese
P
,
Pratt
GA
,
Wang
X
,
Wei
X
,
Xiao
R
, et al
.
A large-scale binding and functional map of human RNA-binding proteins
.
Nature
2020
;
583
:
711
9
.
35.
Yamauchi
T
,
Nishiyama
M
,
Moroishi
T
,
Yumimoto
K
,
Nakayama
KI
.
MDM2 mediates nonproteolytic polyubiquitylation of the DEAD-Box RNA helicase DDX24
.
Mol Cell Biol
2014
;
34
;
3321
40
.
36.
Chao
CH
,
Chen
CM
,
Cheng
PL
,
Shih
JW
,
Tsou
AP
,
Lee
YH
.
DDX3, a DEAD box RNA helicase with tumor growth-suppressive property and transcriptional regulation activity of the p21waf1/cip1 promoter, is a candidate tumor suppressor
.
Cancer Res
2006
;
66
;
6579
88
.
37.
Chang
PC
,
Chi
CW
,
Chau
GY
,
Li
FY
,
Tsai
YH
,
Wu
JC
, et al
.
DDX3, a DEAD box RNA helicase, is deregulated in hepatitis virus-associated hepatocellular carcinoma and is involved in cell growth control
.
Oncogene
2006
;
25
;
1991
2003
.
38.
Xue
Y
,
Jia
X
,
Li
C
,
Zhang
K
,
Li
L
,
Wu
J
, et al
.
DDX17 promotes hepatocellular carcinoma progression via inhibiting Klf4 transcriptional activity
.
Cell Death Dis
2019
;
10
;
814
.
39.
Zhang
T
,
Ma
Z
,
Liu
L
,
Sun
J
,
Tang
H
,
Zhang
B
, et al
.
DDX39 promotes hepatocellular carcinoma growth and metastasis through activating Wnt/β-catenin pathway
.
Cell Death Dis
2018
;
9
;
675
.
40.
Verrou
KM
,
Galliou
PA
,
Papaioannou
M
,
Koliakos
G
.
Phosphorylation mapping of Laminin β1-chain: Kinases in association with active sites
.
J Biosci
2019
;
44:55
.
41.
Chen
Z
,
Xu
X
.
Roles of nucleolin. Focus on cancer and anti-cancer therapy
.
Saudi Med J
2016
;
37
;
1312
8
.
42.
Lee
JH
,
Lee
YS
,
Jeong
SA
,
Khadka
P
,
Roth
J
,
Chung
IK
.
Catalytically active telomerase holoenzyme is assembled in the dense fibrillar component of the nucleolus during S phase
.
Histochem Cell Biol
2014
;
141
;
137
52
.
43.
Daizumoto
K
,
Yoshimaru
T
,
Matsushita
Y
,
Fukawa
T
,
Uehara
H
,
Ono
M
, et al
.
A DDX31/mutant-p53/EGFR axis promotes multistep progression of muscle-invasive bladder cancer
.
Cancer Res
2018
;
78
;
2233
47
.
44.
Fong
KW
,
Li
Y
,
Wang
W
,
Ma
W
,
Li
K
,
Qi
RZ
, et al
.
Whole-genome screening identifies proteins localized to distinct nuclear bodies
.
J Cell Biol
2013
;
203
;
149
64
.
45.
Guo
X
,
Xiong
L
,
Yu
L
,
Li
R
,
Wang
Z
,
Ren
B
, et al
.
Increased level of nucleolin confers to aggressive tumor progression and poor prognosis in patients with hepatocellular carcinoma after hepatectomy
.
Diagn Pathol
2014
;
9
;
175
.
46.
Nguyen
TA
,
Jones
RD
,
Snavely
AR
,
Pfenning
AR
,
Kirchner
R
,
Hemberg
M
, et al
.
High-throughput functional comparison of promoter and enhancer activities
.
Genome Res
2016
;
26
;
1023
33
.
47.
Garg
A
,
Futcher
B
,
Leatherwood
J
.
A new transcription factor for mitosis: in Schizosaccharomyces pombe, the RFX transcription factor Sak1 works with forkhead factors to regulate mitotic expression
.
Nucleic Acids Res
2015
;
43
;
6874
88
.

Supplementary data