CRISPR screening reveals that RNA helicase DDX41 triggers ribosome biogenesis and cancer progression through R-loop-mediated RPL/RPS transcription

Abstract

The RNA helicase DDX41 is a DEAD-box helicase that is well known as a virus sensor in dendritic cells and a tumor suppressor that is frequently mutated in myeloid neoplasms. However, the functions and relevance of DDX41 in solid tumors remain largely unexplored. In this study, through in vivo CRISPR screening, we demonstrate that DDX41 is highly expressed in various solid tumor types and promotes tumorigenicity in liver cancer. Mechanistically, DDX41 facilitates R-loop processing and accelerates the transcription of RPL/RPS genes, thereby promoting ribosome biogenesis and protein synthesis. Additionally, we show that the acetyltransferase KAT8 is required for H3K9ac modification of the DDX41 promoter and that NR2C1/NR2C2 are responsible for DDX41 expression. Moreover, elevated DDX41 levels increase liver cancer cell sensitivity to protein synthesis inhibitors; treatment with homoharringtonine (HHT), an approved drug, significantly inhibits tumor growth in DDX41-overexpressing liver cancer models. Taken together, the results of this study highlight that DDX41 acts as an oncogene in liver cancer and suggest that protein synthesis inhibition may be a promising therapy for liver cancers with high DDX41 expression.

Fig. 1. In vivo CRISPR screening reveals that DDX41 enhances the tumorigenicity of HCC.

a Venn diagram showing the overlap of differentially expressed transcription factors (TFs) and chromosomal regulators (CRs) between 12 paired HCC patients and 50 paired patients from the TCGA-LIHC dataset. b Schematic diagram illustrating the CRISPR screening process. c Volcano plot displaying downregulated (Down) and upregulated (Up) genes identified in the CRISPR screening. Non-targeting control sgRNAs showed as gray dots. d Genes ranked by their corresponding RIGER p-values from the CRISPR screening, analyzed using the RIGER algorithm. e DDX41 mRNA expression levels in normal tissues (NT) and paired tumor tissues (T) from the GSE77314, TCGA-LIHC 50-paired, and GSE77314 cohorts, analyzed using a two-tailed paired Student’s t-test (n = 50 normal tissues and paired tumor tissues). f Violin plot showing DDX41 mRNA expression in normal tissues (NT, n = 50) and tumor tissues (T, n = 105) from the GepLiver database, analyzed with a two-tailed unpaired Student’s t-test. g Pan-cancer analysis of DDX41 mRNA expression in normal tissues (N) and tumor tissues (T), with tumor types marked in red indicating significant differences between N and T; analysis performed using the GEPIA database (The number of samples in each group is indicated above). h Pan-cancer analysis of DDX41 protein expression in normal tissues (Normal) and tumor tissues (Tumor), conducted using the UALCAN database. The upper edge and lower edge of error bar represent the maxima and minima respectively, the upper edge and lower edge of box represent the upper and lower quartile respectively, the line in the box represent the median (The number of samples in each group is indicated above). i Scatter plot showing the correlation between DDX41 and MKI67 mRNA expression levels in the TCGA-LIHC cohort, analyzed using the GEPIA database. Source data are provided as a Source Data file.

Fig. 2. DDX41 promotes liver cancer cell proliferation and is essential for liver cancer initiation.

a Western blot analysis measuring DDX41 protein levels in HUH7 and HEP1 cells infected with negative control (NC) or two independent DDX41 shRNA lentiviruses. ACTIN was used as a loading control. b CCK8 assay assessing the proliferation rate of HUH7 and HEP1 cells infected with NC or two independent DDX41 shRNA lentiviruses. Data are presented as means ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. c Colony formation assay measuring the colony formation ability of HUH7 and HEP1 cells after infection with NC or two independent DDX41 shRNA lentiviruses. Colony formation statistics are shown on the right. Data are presented as means ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. d Western blot analysis measuring DDX41 protein levels in two independent DDX41 double knockout clones of HUH7 cells and DDX41 overexpressed (41-OE) in double knockout clones, with the sgNC clone of HUH7 cells as a negative control. ACTIN was used as a loading control. e CCK8 assay assessing the proliferation rate in the 9# and 20# DDX41 double knockout HUH7 clones or DDX41 overexpressed (41-OE) in the 9# and 20# clones, with sgNC clone of HUH7 cells as a negative control. Data are presented as means ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. f Colony formation assay measuring the colony formation ability of the 9# and 20# DDX41 double knockout HUH7 clones or DDX41 overexpressed (41-OE) in the 9# and 20# clones, with sgNC clone of HUH7 cells as a negative control. Colony formation statistics are shown on the right. Data are presented as means ± SD (n = 3 independent experiments), analyzed using one-way ANOVA with Tukey’s multiple comparisons test. g, h HUH7 cells were infected with DDX41 shRNA lentivirus (shDDX41) to establish DDX41 knockdown clones. HUH7 cells were subcutaneously transplanted into nude mice (n = 6). g Statistics of xenograft tumor weight. Data are presented as means ± SD, analyzed with two-tailed unpaired Student’s t-test. h Statistics of tumor growth curves are presented, with images of xenograft tumors shown above (scale bar: 2 cm). Data are presented as means ± SD, analyzed using two-way ANOVA with Tukey’s multiple comparisons test. i H&E staining and immunohistochemistry (IHC) analysis of DDX41 and Ki67 expression levels in xenograft tumor samples. j Representative images of livers dissected from MYC+β-Catenin and MYC+β-Catenin+sgDdx41 hydrodynamic injection (HDI)-induced mouse liver cancer models. The table presents the number of deaths, tumors, and tumor-free mice in the two groups. k Liver weight, number of visible tumors, and greatest tumor diameter in the livers of MYC+β-Catenin and MYC+β-Catenin+sgDdx41 HDI-induced mouse liver cancer models. Data are presented as means ± SD (n = 3 in group of MYC+β-Catenin; n = 5 in group of MYC+β-Catenin+sgDdx41), analyzed with two-tailed unpaired Student’s t-test. l H&E staining and IHC staining of liver tissues to analyze the expression of MYC, β-Catenin, Ddx41, and Ki67 in HDI-induced mouse liver cancer models. Scale bar: 50 μm (except as indicated). Experiments were repeated three times independently with similar results (a, d). Source data are provided as a Source Data file.

Fig. 3. DDX41 promotes ribosome biogenesis and protein synthesis by increasing RPL/RPS expression in liver cancer cells.

a Venn diagram illustrating the overlap of downregulated genes from RNA-seq results of HUH7 cells transfected with two independent DDX41 siRNAs versus NC (Log₂(FC) < −0.5). A total of 842 genes were downregulated in both DDX41 siRNA-transfected HUH7 cells. b Scatterplot showing the common GSEA negatively regulated GO-BP pathways (n = 526) in RNA-seq data of HUH7 cells following DDX41 knockdown compared to negative control (NC). c GO-BP analysis of the 842 downregulated genes conducted using the DAVID online tool, with the number of genes indicated in the columns. d Heatmap displaying RPL/RPS mRNA expression among the 842 downregulated genes from RNA-seq results, with ACTB as a control. e Western blot analysis measuring RPL/RPS protein levels in HUH7 and HEP1 cells after DDX41 knockdown via two independent DDX41 shRNA or negative control (Ctrl) lentivirus. ACTIN was used as a loading control. f, g Bubble diagrams showing the correlation between DDX41 expression and RPL/RPS levels in three different human liver cancer cohorts (f) or HDI-induced liver cancer mouse models (g). h, i Ribosomal subunit profiles following puromycin-mediated dissociation of HUH7 (h) or HEP1 (i) cells after DDX41 knockout. j, k Polysome profiles of HUH7 (j) or HEP1 (k) cells after DDX41 knockout. l, m SUnSET assay measuring protein synthesis rates in HUH7 cells following DDX41 knockdown via two independent shDDX41 lentiviruses (l) or in DDX41 knockout and DDX41 overexpression conditions (m). ACTIN was detected as a loading control. n OP-Puro assay measuring protein synthesis rates in HUH7 and HEP1 cells after DDX41 knockdown using two independent shDDX41 lentiviruses. Scale bar: 100 μm. o, p SUnSET assay (o) and OP-Puro assay (p) measuring protein synthesis rates in DDX41 high-expressing cell lines (HUH7, HEP1) and low-expressing cell lines (97H, 97 L), with ACTIN as a loading control. Scale bar: 100 μm. Experiments were repeated three times independently with similar results (e, l–p). Source data are provided as a Source Data file.

Fig. 4. DDX41 increases RPL/RPS expression by process the R-loop structures of RPL/RPS genes in liver cancer cells.

a Western blot analysis of DDX41 subcellular localization in HUH7 and HEP1 cells following nuclear-cytoplasmic separation. b Immunofluorescent (IF) staining to analyze DDX41 subcellular localization in HUH7 and HEP1 cells. Scale bar: 25 μm. c Co-immunoprecipitation (co-IP) assay detecting DDX41-interacting proteins identified in the co-IP/MS analysis. d Co-immunofluorescent (co-IF) staining to visualize the co-localization of DDX41 and TOP1 (an R-loop marker) in HUH7 and HEP1 cells. Scale bar: 25 μm. e Co-immunofluorescent (co-IF) staining to detect the co-localization of DDX41 and R-loop structures (using S9.6 antibody to detect DNA hybrids) in HUH7 cells, with RNaseH1 treatment as a negative control. Scale bar: 25 μm. f Correlation between DDX41 occupancy and R-ChIP signals, with reads from HUH7 cells mapped at each ChIP-seq peak and normalized to reads per million. g Upper panel: Signal intensity profiles of ChIP-seq signals for DDX41 and R-ChIP mapped R-loops within ± 3 kb of R-loop centers in HUH7 cells. Lower panel: Heatmap presentation of ChIP-seq signals for DDX41 and R-ChIP mapped R-loops in the same regions. h Genomic distribution of R-ChIP mapped R-loops and DDX41 occupancy. i Sequence features associated with R-loops mapped with R-ChIP and DDX41 ChIP-seq. j DDX41 ChIP-seq (ChIP-41) and R-loop ChIP-seq (R-ChIP) assays illustrating the peaks for DDX41 and R-loop structures on RPL/RPS genes in HUH7 cells. k R-ChIP‒qPCR and DDX41 ChIP‒qPCR showing the enrichment (relative to 1% input) of R-loop structures and DDX41 on RPL/RPS genes in HUH7 cells. Data are presented as mean ± SD (n = 3 independent experiments). l, m Dot blot analysis measuring R-loop structure levels in HUH7 cells following DDX41 knockout (l) or DDX41 overexpression (m), with sgNC or vector as negative controls. Methylene blue staining was used as a loading control. n R-ChIP‒qPCR measuring the relative enrichment (1% input) of R-loop structures at RPL/RPS genes in HUH7 cells after DDX41 knockout, with sgNC as a negative control. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. o R-ChIP‒qPCR measuring the relative enrichment (1% input) of R-loop structures at RPL/RPS genes in DDX41 knockout cells (9#) or DDX41 knockout cells rescued DDX41 expression (41-OE), with sgNC as a negative control. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. p Real-time qPCR measuring the expression of RPL/RPS nascent mRNAs in DDX41 knockout cells (9#) or DDX41 knockout cells rescued DDX41 expression (41-OE), with sgNC as a negative control. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. Experiments were repeated three times independently with similar results (a–e, l and m). Source data are provided as a Source Data file.

Fig. 5. The DEAD-box/HCD/ZnF domain of DDX41 is essential for its ability to promote RPL/RPS expression and R-loop structure localization.

a Schematic diagram illustrating the four domains of the DDX41 protein and the truncations we constructed. b Immunofluorescent (IF) staining showing the subcellular localization of DDX41 and DDX41 with deleted nuclear localization sequences (NLSs) in HUH7 cells. Scale bar: 25 μm. c Immunofluorescent (IF) staining depicting the subcellular localization of DDX41 with various deleted domains. Scale bar: 25 μm. d Immunofluorescent (IF) staining illustrating the subcellular localization of DDX41 with only the HCD domain or with deleted NLSs and HCD domain. Scale bar: 25 μm. e Scatterplot showing proteins (n = 84) detected only in Flag-tagged HCD (Flag-HCD) compared to IgG control. The X-axis represents coverage percentage, and the Y-axis represents -Log₂(P value). f Co-immunoprecipitation (Co-IP) assay and western blot analysis measuring the interaction between DDX41, KPNA2, and KPNB1. g Real-time qPCR measuring the expression of RPL/RPS mRNAs in DDX41 knockout HUH7 cells (9#) overexpressing DDX41 truncations. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. h Flag-tagged ChIP-qPCR measuring the relative enrichment (1% input) of DDX41 truncations on RPL/RPS genes in DDX41 knockout HUH7 cells (9#). Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. Experiments were repeated three times independently with similar results (b–d and f). Source data are provided as a Source Data file.

Fig. 6. KAT8 promotes H3K9ac modification of the DDX41 promoter and recruits NR2C1/2 to induce DDX41 expression in liver cancer cells.

a ChIP-seq analysis for H3K9ac and H3K27ac showing the peaks of H3K9ac and H3K27ac on the DDX41 promoter in liver cell lines and normal liver tissue, with H3K4me3 peaks indicating the promoter region. b Real-time qPCR demonstrating DDX41 mRNA expression (relative to ACTB) in HUH7 and HEP1 cells after transfection with various acetyltransferase siRNAs, with NC as the negative control. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using one-way ANOVA with Tukey’s multiple comparisons test. c Western blot analysis measuring DDX41 and KAT8 protein expression in HUH7 and HEP1 cells after transfection with two independent KAT8 siRNAs (si8-1/si8-2), with NC as the negative control. ACTIN was used as a loading control. d ChIP-seq data showing the peaks of H3K9ac and KAT8 on the DDX41 promoter. e ChIP‒qPCR of H3K9ac demonstrating relative enrichment (1% input) on the DDX41 promoter in HUH7 cells after KAT8 knockdown. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. f Luciferase assay measuring the active core region of the DDX41 promoter in HUH7 cells using a series of truncated promoter regions. Schematic diagram on the left. TSS: transcription start site; RLU: relative luciferase unit. Data are presented as mean ± SD (n = 3 independent experiments). g Correlation analysis between DDX41 expression and predicted transcription factor candidates in the TCGA-LIHC cohort from the GEPIA database. The X-axis represents -Log10(P value), and numbers in the columns indicate the R value. h Luciferase assay assessing DDX41 promoter relative activities in HUH7 cells after knockdown of NR2C1 and/or NR2C2. RLU: relative luciferase unit. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using one-way ANOVA with Tukey’s multiple comparisons test. i Western blot measuring DDX41, NR2C1, and NR2C2 protein expression in HUH7 and HEP1 cells after NR2C1 and/or NR2C2 knockdown, with NC as the negative control. ACTIN was detected as a loading control. j ChIP-seq data showing the peaks of NR2C1 and NR2C2 at the DDX41 promoter in HUH7 cells. k ChIP‒qPCR measuring the relative enrichment (1% input) of NR2C1 (Flag-C1) and NR2C2 (Flag-C2) on the DDX41 promoter in HUH7 cells. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using one-way ANOVA with Tukey’s multiple comparisons test. l ChIP‒qPCR measuring the relative enrichment (1% input) of NR2C1 (Flag-C1) and NR2C2 (Flag-C2) on the DDX41 promoter in HUH7 cells after KAT8 knockdown. Data are presented as mean ± SD (n = 3 independent experiments), analyzed using two-way ANOVA with Tukey’s multiple comparisons test. Experiments were repeated three times independently with similar results (c and l). Source data are provided as a Source Data file.

Fig. 7. Protein synthesis inhibitor treatment dramatically restricts tumor growth and progression in DDX41-overexpressing liver cancer cells in vivo.

a CCK8 assay measuring cell viability of HUH7 and HEP1 cells after DDX41 overexpression and treatment with various concentrations of Homoharringtonine (HHT), with vector as control. Data are presented as mean ± SD (n = 3 independent experiments). b CCK8 assay measuring cell viability of HUH7 and HEP1 cells after DDX41 knockout and treatment with various concentrations of HHT, with sgNC as control. Data are presented as mean ± SD (n = 3 independent experiments). c, d Colony formation assay evaluating colony formation ability of HUH7 (c) and HEP1 (d) cells after DDX41 overexpression and treatment with various concentrations of HHT, with vector as control. Colony formation statistics are shown on the right, data are presented as mean ± SD (n = 3 independent experiments). Two-way ANOVA with Tukey’s multiple comparisons test. e, f Colony formation assay assessing colony formation ability of HUH7 (e) and HEP1 (f) cells with or without DDX41 knockout and treated with various concentrations of HHT, with sgNC as control. Colony formation statistics are shown on the right, data are presented as mean ± SD (n = 3 independent experiments). Two-way ANOVA with Tukey’s multiple comparisons test. g–j HUH7 cells stably overexpressing DDX41 (vector (V) as control) were subcutaneously transplanted into nude mice (n = 6). Mice were treated with HHT (5 mg/kg) or vehicle every two days starting 8 days post-transplantation. Images of xenograft tumors are shown in (g). Tumor volumes were measured every 2–3 days, with tumor growth curves presented in (h), data are presented as mean ± SD, analyzed with Two-way ANOVA with Tukey’s multiple comparisons test. Tumor weights in (i), data are presented as mean ± SD, the upper edge and lower edge of error bar represent the maxima and minima respectively, the upper edge and lower edge of box represent the upper and lower quartile respectively, the line in the box represent the median. Analyzed with Two-way ANOVA with Tukey’s multiple comparisons test. Relative tumor growth inhibition ratios in (j), data are presented as mean ± SD, analyzed with Two-way ANOVA with Tukey’s multiple comparisons test. k H&E staining and immunohistochemistry (IHC) to analyze DDX41 and Ki67 expression in xenograft tumor samples. Scale bars: 200 μm for H&E staining and DDX41/ki67 IHC; 100 μm for enlarged Ki67 IHC. l Proposed working model illustrating the mechanisms by which high DDX41 expression promotes RPL/RPS expression and protein synthesis in liver cancer cells. Source data are provided as a Source Data file.