The pleiotropic roles of DEAD-box helicase 3, X-linked (DDX3X), including its functions in transcriptional and translational regulation, chromosome segregation, DNA damage, and cell growth control, have highlighted the association between DDX3X and tumorigenesis. However, mRNA transcripts and protein levels of DDX3X in patient specimens have shown the controversial correlations of DDX3X with hepatocellular carcinoma (HCC) prevalence. In this study, generation of hepatocyte-specific Ddx3x-knockout mice revealed that loss of Ddx3x facilitates liver tumorigenesis. Loss of Ddx3x led to profound ductular reactions, cell apoptosis, and compensatory proliferation in female mutants at 6 weeks of age. The sustained phosphorylation of histone H2AX (γH2AX) and significant accumulation of DNA single-strand breaks and double-strand breaks in liver indicated that the replicative stress occurred in female mutants. Further chromatin immunoprecipitation analyses demonstrated that DDX3X bound to promoter regions and regulated the expression of DNA repair factors, DDB2 and XPA, to maintain genome stability. Loss of Ddx3x led to decreased levels of DNA repair factors, which contributed to an accumulation of unrepaired DNA damage, replication stress, and eventually, spontaneous liver tumors and DEN-induced HCCs in Alb-Cre/+;Ddx3xflox/flox mice.

Implications:

These data identify an important role of DDX3X in the regulation of DNA damage repair to protect against replication stress in liver and HCC development and progression.

Liver cancer is the fifth most common cancer and the third most frequent cancer-related of deaths worldwide. The majority of liver malignancies are hepatocellular carcinoma (HCC) and cholangiocarcinoma that have high morbidity and mortality in developing countries, but their incidences are also increased in developed countries. Etiological studies of liver cancer have identified four major risk factors, including chronic infection with hepatitis B or C viruses, alcohol use, and the intake of aflatoxin-contaminated food (1–3). In recent years, obesity-associated nonalcoholic fatty liver disease has continuously increased (4). It is generally accepted that hepatocarcinogenesis is a multistep process with an accumulation of the genetic and epigenetic alterations of critical genes involving hepatocyte growth and proliferation, cell death, cell adhesion, and metabolism (5). A common feature during HCC progression is repeated cycles of cell death and compensatory proliferation, which causes an inflammatory microenvironment in the liver, leading to fibrosis, cirrhosis, and ultimately, malignant transformation of the hepatocytes.

DEAD-box helicase 3, X-linked (DDX3X; also known as DDX3) belongs to the DEAD-box RNA helicase family and was initially identified as a cellular factor that bound to hepatitis C viral core protein. Accumulated evidences showed that DDX3X is involved in diverse cellular processes, including cell growth, cell-cycle control, mitotic chromosome segregation, innate immunity, tumorigenesis, and virus replication (6). Our previous studies showed that DDX3X acted as a tumor suppressor in various cancer cell lines. We found that lower DDX3X levels were significantly correlated with higher HCC prevalence, particularly in male and HBV-positive patients (7). Knockdown of DDX3X in nontransformed NIH3T3 cells led to a premature entry into S phase and an increase of cell proliferation, and enhanced the oncogenic v-ras–induced anchorage-independent growth. In addition, overexpression of DDX3X activated the p21WAF1/Cip1 promoter through its physical interaction with Sp1, and consequently, negatively regulated cyclin D1 protein levels and tumor cell proliferation (8). A recent study further showed that DDX3X level was inversely associated with the tumor grade and predicted poor prognoses for HCC patients. Epigenetic regulation of a subset of HCC-associated tumor-suppressor miRNAs, including miR-200b, miR-200c, miR-122, and miR-145, by DDX3X induced stem cell–like properties and promoted tumorigenic potential (9). In contrast to our studies, Huang and colleagues showed by qRT-PCR analysis that DDX3X was overexpressed in 64% of HCC tissues (10), and Su and colleagues found no association between DDX3X RNA level and survival in liver cancer patients (11). Therefore, the pathologic role of DDX3X in liver cancer remains to be clarified. In a previous study, we generated a conditional knockout allele for Ddx3x gene and identified the essential roles of DDX3X in placental and early embryonic development (12). Here, hepatocyte-specific Ddx3x-knockout mice were generated. We showed that loss of Ddx3x led to DNA damage and replicative stress at a young age, and mice developed spontaneous liver tumors with aging. Ddx3x-deficiency females were more susceptible to diethylnitrosamine (DEN)-induced HCC than littermate female controls. Moreover, DDX3X regulated DNA damage repair factors DDB2 and XPA, in the liver and in vitro, which are critical to maintain genome stability.

Mice, DEN treatment, and serum biochemical analysis

The floxed Ddx3x mice were generated previously (12) and had been backcrossed into the C57BL/6 background for at least 10 generations. The hepatocyte-specific Ddx3x-knockout mice were generated through breeding Ddx3xflox/flox mice with Alb-Cre [B6.Cg-Tg(Alb-cre)21Mgn/J] transgenic mice (13). The serum levels of alanine aminotransferase (ALT), triacylglycerol (TG), and total cholesterol (TCHO) were monitored using biochemical slides (Fuji DRY-CHEM 400i; Fujifilm) according to the manufacturer's instructions. To induce HCC, DEN (25 mg/kg body weight; N0756-10ML, Sigma-Aldrich) was injected i.p. into 2-week-old mice. To induce acute liver injury, 10-week-old mice were injected with DEN (100 mg/kg body weight) and sacrificed 1 or 3 days thereafter. The experimental procedures using mice were approved by the Institutional Animal Care and Use Committee (IACUC) of National Yang-Ming University, Taiwan. The animal care and experimental procedures were performed in accordance with the Guidelines of the IACUC of National Yang-Ming University, Taiwan.

Histologic analysis and molecular techniques

Further details are provided in the Supplementary Information.

Statistical analysis

Hepatocytes stained positive for different markers in livers were determined by the positive hepatocytes/total hepatocytes in each high-power (200× magnification) field. For each sample, at least four fields were randomly selected and counted. Quantitative results are expressed as the mean ± SEM. The Student t test was used to determine P values, unless stated otherwise.

Hepatocyte-specific Ddx3x ablation leads to the development of hepatocellular tumors in aged mice

To investigate the physiopathologic role of DDX3X in liver, floxed Ddx3x mice were crossed with Alb-Cre transgenic mice. The Alb-Cre transgene is specifically activated in the liver from late embryonic stage, and Cre-mediated gene recombination occurs progressively with age (14, 15). PCR of genomic DNA from liver confirmed the progressive loss of floxed Ddx3x allele and the presence of null allele. The reduction of DDX3X protein was specifically observed in liver. The presence of nonparenchymal cells, which do not have Cre activity, may explain the low levels of DDX3X protein in livers of mutant mice (Supplementary Fig. S1). The hepatocyte-specific Ddx3x-knockout mice were born at expected Mendelian ratios and were morphologically indistinguishable from their littermate controls at young ages. Therefore, serum ALT, TG, and TCHO levels, commonly used markers for monitoring liver injury and functions, were measured routinely every 3 months from 12 months of age. TG and TCHO levels in the mutant mice were not different from their littermate controls at all time periods (Supplementary Fig. S2). However, ALT levels were elevated in some, but not all, of the hepatocyte-specific Ddx3x-knockout mice with visible tumors, when dissected (Fig. 1A). The results showed that 10.7% (3/28) and 42.1% (8/19) of dissected Alb-Cre/+;Ddx3xflox/Y male and Alb-Cre/+;Ddx3xflox/flox female mutants, respectively, developed liver tumors after approximately 12 to 27 months. Among these mice, we noted 4 of 5 (80%) Alb-Cre/+;Ddx3xflox/flox females developed liver tumors at 24 months of age. None of the littermate controls and Alb-Cre/+;Ddx3xflox/+ heterozygous females had a tumor (Fig. 1B and C). Hematoxylin & eosin (H&E) staining and immunohistochemical analysis of hepatocyte marker HNF4α and cholangiocyte markers K19 and Sox9 in tumor sections revealed that the most common histopathology was the clear cell type HCC (16), which was detected in both Alb-Cre/+;Ddx3xflox/flox female and Alb-Cre/+;Ddx3xflox/Y male. Other types of liver tumors, including adenoma and cholangiocarcinoma, were observed in Alb-Cre/+;Ddx3xflox/flox female and Alb-Cre/+;Ddx3xflox/Y male, respectively (Fig. 1D). Western blot analysis confirmed that DDX3X protein levels were significantly decreased in tumor and nontumor liver tissues from aged Alb-Cre/+;Ddx3xflox/Y male and Alb-Cre/+;Ddx3xflox/flox female mice compared with their gender-matched littermate controls (Fig. 1E). Given the higher tumor incidence was observed in aged Alb-Cre/+;Ddx3xflox/flox females, the effects of DDX3X on liver tumorigenesis in female mice were extensively characterized.

Figure 1.

Spontaneous development of liver tumors in hepatocyte-specific Ddx3x knockout mice. The sera and livers were collected from Ddx3x-knockout mice (Alb-Cre/+;Ddx3xflox/Y males and Alb-Cre/+;Ddx3xflox/flox females) and their littermate controls (Ddx3x+/Y, Alb-Cre/+;Ddx3x+/Y, or Ddx3xflox/Y males and Ddx3xflox/+ or Ddx3xflox/flox females) from 12 to 27 months of age. A, The serum ALT levels. N ≥ 15 per group. B, Tumor-free survival of controls and Ddx3x-knockout mice was assessed using Kaplan–Meier analysis. C, The numbers and sizes of tumors in mice. The size of the liver tumor was analyzed by determining the longest diameter of a tumor. D, Gross images of representative livers and H&E and IHC staining of liver sections at indicated ages (M, month). The cholangiocellular carcinoma (CC) and clear cell type HCC were observed in Alb-Cre/+;Ddx3xflox/Y males. The clear cell type HCC was found in the Alb-Cre/+;Ddx3xflox/flox females. Expression of hepatocyte marker HNF4α and cholangiocyte markers K19 and SOX9 in tumors of Ddx3x-knockout mice was assessed by IHC. Dashed lines demarcate the boundary between normal liver tissue (N) and tumor (T). E, Expression levels of DDX3X in liver tissues from controls and the tumors (T) and adjacent liver tissues (N) from Ddx3x-knockout mice were evaluated by Western blot analysis. Protein levels were normalized to the loading control GAPDH, expressed as fold changes relative to gender-matched controls (set as 1) and shown under their corresponding panels. N = 4 per group. One-way ANOVA was used to determine P values in A and C. Log-rank (Mantel–Cox) test was used to determine P values in B. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bars, 1 cm (gross view); 200 μm (section).

Figure 1.

Spontaneous development of liver tumors in hepatocyte-specific Ddx3x knockout mice. The sera and livers were collected from Ddx3x-knockout mice (Alb-Cre/+;Ddx3xflox/Y males and Alb-Cre/+;Ddx3xflox/flox females) and their littermate controls (Ddx3x+/Y, Alb-Cre/+;Ddx3x+/Y, or Ddx3xflox/Y males and Ddx3xflox/+ or Ddx3xflox/flox females) from 12 to 27 months of age. A, The serum ALT levels. N ≥ 15 per group. B, Tumor-free survival of controls and Ddx3x-knockout mice was assessed using Kaplan–Meier analysis. C, The numbers and sizes of tumors in mice. The size of the liver tumor was analyzed by determining the longest diameter of a tumor. D, Gross images of representative livers and H&E and IHC staining of liver sections at indicated ages (M, month). The cholangiocellular carcinoma (CC) and clear cell type HCC were observed in Alb-Cre/+;Ddx3xflox/Y males. The clear cell type HCC was found in the Alb-Cre/+;Ddx3xflox/flox females. Expression of hepatocyte marker HNF4α and cholangiocyte markers K19 and SOX9 in tumors of Ddx3x-knockout mice was assessed by IHC. Dashed lines demarcate the boundary between normal liver tissue (N) and tumor (T). E, Expression levels of DDX3X in liver tissues from controls and the tumors (T) and adjacent liver tissues (N) from Ddx3x-knockout mice were evaluated by Western blot analysis. Protein levels were normalized to the loading control GAPDH, expressed as fold changes relative to gender-matched controls (set as 1) and shown under their corresponding panels. N = 4 per group. One-way ANOVA was used to determine P values in A and C. Log-rank (Mantel–Cox) test was used to determine P values in B. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bars, 1 cm (gross view); 200 μm (section).

Close modal

Hepatocyte-specific deletion of Ddx3x leads to liver injury, ductular reactions, and inflammation

To further determine the pathologic role of DDX3X in tumor progression, serum ALT levels were closely monitored at young ages. The mean ALT level of 3- to 20-week-old littermate controls was 13.4 ± 0.7 U/l (Fig. 2A). We found significantly elevated ALT levels in Alb-Cre/+;Ddx3xflox/flox mutants at 6 and 10 weeks of age (115.1 ± 12.0 and 26.9 ± 2.0 U/l, respectively). There was no difference in ALT levels between controls and mutants at 3 and 20 weeks of age. Histologic analyses of H&E-stained liver sections revealed increased numbers of basophilic cells [i.e., ductular reactions (DRs)] in 6- and 10-week-old Alb-Cre/+;Ddx3xflox/flox mice, in accordance with increased serum ALT levels, when compared with controls (Fig. 2B). The DRs (17) were further verified by significant expansion of liver progenitor cell (LPC)/cholangiocyte marker–positive subsets, including A6-, EpCAM-, cytokeratin 19 (K19)-, and Sox9-positive cells as assessed by IHC in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Fig. 2C). DRs were notably reduced to a lesser extent in 10-week-old Alb-Cre/+;Ddx3xflox/flox livers and disappeared thereafter (Supplementary Fig. S3). qRT-PCR analyses confirmed a decrease in Ddx3x expression and significant increases of LPC/cholangiocyte gene levels, including CD133, EpCAM, and K19 in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers compared with those of littermate controls. Nevertheless, the increasing trend of Sox9 levels did not reach statistical significance (Fig. 2D). The liver injury was also associated with significant immune cell infiltration consisting of F4/80 (macrophages)-, CD3 (T cells)-, and B220 (B cells)-positive cells and increased Tnfα levels (Supplementary Fig. S4), which coincided with the significant elevation of ALT levels in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers compared with their controls. In aged mice, F4/80-, B220-, and CD3-positive cells were significantly increased in nontumor liver tissues of tumor-bearing Alb-Cre/+;Ddx3xflox/flox mice compared with those of littermate controls. We found that IL1β-, TNFα-, and NF-κB–positive cells were significantly increased in both tumors and adjacent nontumor liver tissues of aged tumor-bearing Alb-Cre/+;Ddx3xflox/flox livers. Also, Il1β and Tnfα mRNA levels were significantly increased in tumors from Alb-Cre/+;Ddx3xflox/flox livers (Supplementary Fig. S5).

Figure 2.

Loss of Ddx3x results in liver injury and DRs in Alb-Cre/+;Ddx3xflox/flox mice. A, The serum ALT activity was measured in controls (Ddx3xflox/+ and Ddx3xflox/flox) and Alb-Cre/+;Ddx3xflox/flox mice at indicated ages (age in weeks). N ≥ 5 per group. B, Representative images of H&E-stained liver sections. DRs (arrowheads) were observed in Alb-Cre/+;Ddx3xflox/flox livers at 6 and 10 weeks of age. C, Representative A6, EpCAM, K19, and Sox9 immunostaining (red) in 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox livers. The positively stained cells are shown at higher magnification in the inset. Nuclei (blue) were counterstained with hematoxylin. D, Relative mRNA levels of Ddx3x and cholangiocyte/LPC genes in livers from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice were detected by qRT-PCR. N = 5 per group. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bar, 200 μm.

Figure 2.

Loss of Ddx3x results in liver injury and DRs in Alb-Cre/+;Ddx3xflox/flox mice. A, The serum ALT activity was measured in controls (Ddx3xflox/+ and Ddx3xflox/flox) and Alb-Cre/+;Ddx3xflox/flox mice at indicated ages (age in weeks). N ≥ 5 per group. B, Representative images of H&E-stained liver sections. DRs (arrowheads) were observed in Alb-Cre/+;Ddx3xflox/flox livers at 6 and 10 weeks of age. C, Representative A6, EpCAM, K19, and Sox9 immunostaining (red) in 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox livers. The positively stained cells are shown at higher magnification in the inset. Nuclei (blue) were counterstained with hematoxylin. D, Relative mRNA levels of Ddx3x and cholangiocyte/LPC genes in livers from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice were detected by qRT-PCR. N = 5 per group. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bar, 200 μm.

Close modal

Loss of Ddx3x causes DNA damage, cell death, and compensatory proliferation in young mice

It has generally been accepted that loss of hepatic mass after injury activates compensatory proliferation, leading to tumor development (18). In accordance with the dynamic changes of ALT levels, large numbers of cleaved Caspase-3 (cCasp3)–positive apoptotic cells were specifically detected in the Alb-Cre/+;Ddx3xflox/flox livers at 6 weeks of age when compared with their littermate controls (Fig. 3A and B). In 3-week-old mice, cell proliferation was still high in the livers, and proliferation marker Ki67-positive cells were comparable in livers from both controls and Alb-Cre/+;Ddx3xflox/flox mice. The decrease in cell proliferation was observed in control livers as development proceeded; nevertheless, we detected a significant increase of Ki67-positive cells in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Fig. 3A and C). These results indicated that compensatory proliferation occurred after cell apoptosis in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers. Our previous study showed that targeted Ddx3x ablation in the epiblast leads to widespread γH2AX (phosphorylated histone H2AX, as a marker of DNA damage) and apoptosis, which causes embryonic lethality (12). We noted that γH2AX signals were significantly increased in Alb-Cre/+;Ddx3xflox/flox livers, peaked at 6 weeks of age, and then gradually declined at 10 and 20 weeks of age, compared with littermate controls (Fig. 3A and D). The γH2AX foci were observed in approximately 10% of proliferating Ki67-positive hepatocytes in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Supplementary Fig. S6). The significant higher γH2AX+/Ki67+ hepatocytes in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers imply that loss of Ddx3x causes prolonged DNA damage, and may be linked to genome instability and subsequent liver tumorigenesis in Alb-Cre/+;Ddx3xflox/flox mice.

Figure 3.

Loss of Ddx3x causes cell death, compensatory proliferation, and DNA damage in Alb-Cre/+;Ddx3xflox/flox livers at 6 weeks of age. A, IHC analyses of cCasp3 (red), Ki67 (red), and γH2AX (green) in livers from controls (Ddx3xflox/+ and Ddx3xflox/flox) and Alb-Cre/+;Ddx3xflox/flox mice at indicated ages (age in weeks). Nuclei (blue) were counterstained with hematoxylin and/or 4,6-diamidino-2-phenylindole (DAPI). B–D, Percentages of cCasp3-, Ki67-, and γH2AX-positive hepatocytes in livers from controls and Alb-Cre/+;Ddx3xflox/flox mice were quantified. N ≥ 3 per group. *, P < 0.05 and ***, P < 0.001. Scale bar, 200 μm.

Figure 3.

Loss of Ddx3x causes cell death, compensatory proliferation, and DNA damage in Alb-Cre/+;Ddx3xflox/flox livers at 6 weeks of age. A, IHC analyses of cCasp3 (red), Ki67 (red), and γH2AX (green) in livers from controls (Ddx3xflox/+ and Ddx3xflox/flox) and Alb-Cre/+;Ddx3xflox/flox mice at indicated ages (age in weeks). Nuclei (blue) were counterstained with hematoxylin and/or 4,6-diamidino-2-phenylindole (DAPI). B–D, Percentages of cCasp3-, Ki67-, and γH2AX-positive hepatocytes in livers from controls and Alb-Cre/+;Ddx3xflox/flox mice were quantified. N ≥ 3 per group. *, P < 0.05 and ***, P < 0.001. Scale bar, 200 μm.

Close modal

DNA single-strand break and double-strand break signalings were induced in Alb-Cre/+;Ddx3xflox/flox livers

Proper DNA replication and chromosome segregation are prerequisites for genome integrity. To ensure the normal cellular functions and faithful genome maintenance and transmission, a complex network of DNA damage response (DDR) systems has evolved to sense and respond to different forms of DNA damage and replication stress (19, 20). Along with increased numbers of γH2AX-positive cells, IHC analyses of liver sections revealed that enlarged hepatocytes with atypical nuclei were frequently observed in Alb-Cre/+;Ddx3xflox/flox livers compared with those of control livers (Supplementary Fig. S7A). Notably, the numbers of cells with 53BP1 nuclear bodies were significantly increased in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Supplementary Fig. S7B). These results prompted us to speculate that loss of Ddx3x causes replication stress and may be a driving force for tumorigenesis in Alb-Cre/+;Ddx3xflox/flox livers.

Replication protein A (RPA) is a guardian of genome integrity during DNA replication, recombination, and repair events, and also acts as a key sensor in DDR (21, 22). IHC analyses of 3-week-old liver sections showed similar expression levels of RPA70 and RPA32 in the Alb-Cre/+;Ddx3xflox/flox mice and their littermate controls (Fig. 4A). As development proceeded, the expression levels of both RPA70- and RPA32-positive cells declined significantly in 6- and 10-week-old littermate controls compared with those of 3-week-old controls. The staining intensities of RPA70 and RPA32 persisted in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Fig. 4A). The changes of DDX3X, RPA70, RPA32, and phosphorylated RPA32 (the active form of RPA32) protein levels in livers from 3-, 6-, and 10-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice were further confirmed by Western blotting. As expected, a significant increase of γH2AX expression was detected in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Fig. 4B; Supplementary Fig. S8A).

Figure 4.

DNA SSB and DSB signalings are induced in 6-week-old Alb-Cre/+;Ddx3xflox/flox mice. The liver tissues from controls and Alb-Cre/+;Ddx3xflox/flox mice at 3, 6, and 10 weeks of age were collected and processed for immunostaining and Western blot analysis. A, Representative images of RPA70 and RPA32 (red) immunostaining. B, Representative images of DDX3X, γH2AX, RPA70, and RPA32 protein expression in livers from controls (Ctrl) and Alb-Cre/+;Ddx3xflox/flox (KO) mice by Western blot analysis. The pRPA32 is the phosphorylated form of RPA32. Protein levels were normalized to the loading control GAPDH, expressed as fold changes relative to 3-week-old controls (set as 1) and shown under their corresponding panels. C, Immunofluorescence costaining of pATR (green) and pChk1 (green) with hepatocyte marker HNF4α (red). Percentages of pATR- and pChk1-positive hepatocytes (HNF4α+) in livers were quantified. N ≥ 3 per group. D, Representative images of Mre11 (red) immunostaining. E, Representative images of Mre11 protein expression in livers from controls (Ctrl) and Alb-Cre/+;Ddx3xflox/flox (KO) mice by Western blot analysis. Fold changes of Mre11 expression relative to 3-week-old controls (set as 1) are shown under their corresponding panels. F, Immunofluorescence costaining of pChk2 (green) and HNF4α. Percentages of pChk2-positive hepatocytes in livers were quantified. N = 3 per group. The stained cells (A, C, D, and F) are shown at higher magnification in the inset. Nuclei were stained with hematoxylin (A and D) and/or DAPI (C and F). *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bar, 200 μm.

Figure 4.

DNA SSB and DSB signalings are induced in 6-week-old Alb-Cre/+;Ddx3xflox/flox mice. The liver tissues from controls and Alb-Cre/+;Ddx3xflox/flox mice at 3, 6, and 10 weeks of age were collected and processed for immunostaining and Western blot analysis. A, Representative images of RPA70 and RPA32 (red) immunostaining. B, Representative images of DDX3X, γH2AX, RPA70, and RPA32 protein expression in livers from controls (Ctrl) and Alb-Cre/+;Ddx3xflox/flox (KO) mice by Western blot analysis. The pRPA32 is the phosphorylated form of RPA32. Protein levels were normalized to the loading control GAPDH, expressed as fold changes relative to 3-week-old controls (set as 1) and shown under their corresponding panels. C, Immunofluorescence costaining of pATR (green) and pChk1 (green) with hepatocyte marker HNF4α (red). Percentages of pATR- and pChk1-positive hepatocytes (HNF4α+) in livers were quantified. N ≥ 3 per group. D, Representative images of Mre11 (red) immunostaining. E, Representative images of Mre11 protein expression in livers from controls (Ctrl) and Alb-Cre/+;Ddx3xflox/flox (KO) mice by Western blot analysis. Fold changes of Mre11 expression relative to 3-week-old controls (set as 1) are shown under their corresponding panels. F, Immunofluorescence costaining of pChk2 (green) and HNF4α. Percentages of pChk2-positive hepatocytes in livers were quantified. N = 3 per group. The stained cells (A, C, D, and F) are shown at higher magnification in the inset. Nuclei were stained with hematoxylin (A and D) and/or DAPI (C and F). *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bar, 200 μm.

Close modal

It has been demonstrated that RPA binds to ssDNA and triggers activation of the ATR-Chk1 pathway in response to DNA damage or replication stress (22, 23). Indeed, significant increases of phosphorylated ATR (pATR)– and phosphorylated Chk1 (pChk1)–positive hepatocytes were detected in 6-week-old Alb-Cre/+;Ddx3xflox/flox liver sections compared with those of their littermate controls (Fig. 4C). These results revealed that loss of DDX3X induces replication stress in Alb-Cre/+;Ddx3xflox/flox livers at a young age.

If replication stress persists, ssDNA long-time exposure in cells consequently results in double-strand break (DSB) and apoptosis. The presence of the MRN complex (Mre11, Rad50, and NBS1), the functional DSB sensor, in the DNA damage site, subsequently recruits and activates the ATM-Chk2 signaling pathway (22, 23). IHC staining and Western blot analyses showed increased Mre11 levels in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers compared with their littermate controls (Fig. 4D and E; Supplementary Fig. S8A). In addition, the pChk2-positive hepatocytes were dramatically increased in the 6-week-old Alb-Cre/+;Ddx3xflox/flox livers (Fig. 4F). To substantiate these findings are hepatocyte specific, the expression levels of DDX3X, γH2AX, and the sensors in DDR in isolated hepatocytes from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox livers were examined. In isolated hepatocytes, PCR analysis of genomic DNA showed the floxed Ddx3x alleles were almost completely deleted and DDX3X protein levels from 6-week-old Alb-Cre/+;Ddx3xflox/flox livers were significantly decreased to approximately 4% of control hepatocytes. The relative fold changes of γH2AX, RPA70, pRPA32, RPA32, and Mre11 proteins were greater in isolated hepatocytes than those from whole livers of 6-week-old Alb-Cre/+;Ddx3xflox/flox mice, confirming a hepatocyte-specific effect (Supplementary Fig. S8C compared with Supplementary Fig. S8A). Taken together, these results showed that not only single-strand break (SSB), but also DSB signaling, was induced in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers.

Ddx3x loss inhibits the nucleotide excision repair and downregulates DDB2 and XPA genes through Sp1

Cells halt cell-cycle progression in the presence of DNA damage to provide more time for DNA repair. The levels of Cdkn2b (p15Ink4b), Cdkn1a (p21WAF1/CIP1), and Cdkn1b (p27Kip1) were increased by 2-, 3.8-, and 1.4-fold, respectively, in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers compared with controls, indicating that loss of Ddx3x halted cell-cycle progression in response to DNA damage (Supplementary Fig. S9). Consistent with our previous study in embryonic cells (12), the Trp53 (p53) level was reduced in the Alb-Cre/+;Ddx3xflox/flox livers compared with that of the controls.

The activation of ATR/ATM, upregulation of cell-cycle regulators, and the prolonged γH2AX-positive cells in Alb-Cre/+;Ddx3xflox/flox livers suggest the possibility that the downstream functions of DNA repair and signaling may not work efficiently. The nucleotide excision repair (NER) machinery is involved in the removal of a wide range of DNA lesions either through global genome NER (GG-NER) or through transcription-coupled NER (TC-NER). The damage is initially recognized by DDB2 (damage-specific DNA binding protein 2) and XPC (xeroderma pigmentosum, complementation group C) in GG-NER, or ERCC6 (excision repair cross-complementation group 6) and ERCC8 in TC-NER. After recognition, DDB1 interacts with DDB2 or ERCC8, which then recruits XPA and its interacting proteins to perform DNA unwinding (24). Thereafter, sequential and coordinated assembly of core repair factors functionally involved in DNA incision, excision, and de novo synthesis proceeds through the NER pathway (25). In 6-week-old Alb-Cre/+;Ddx3xflox/flox livers, the expression levels of Ddb2, Ercc6, Ercc8, and Xpa, but not Ddb1 and Xpc, were significantly decreased (Fig. 5A). The expression of Ercc1, Ercc4, Ercc5, and proliferating cell nuclear antigen (Pcna), which are involved in DNA incision, excision, and de novo synthesis, were not affected (Fig. 5A). A survey of the proximal promoter regions identified the confirmed Sp1-binding sites and/or GC-rich elements in the responsive genes, Ddb2 (26), Ercc6 (27), and Xpa (28). We then found that the expression level of Sp1, a previously identified partner of DDX3X, was decreased in 6-week-old Alb-Cre/+;Ddx3xflox/flox livers.

Figure 5.

The targets of DDX3X are involved in NER. A, qRT-PCR analyses of NER genes and Sp1 in livers from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice. N = 4 per group. B, Relative mRNA levels of DDX3X, DDB2, ERCC6, ERCC8, XPA, and Sp1 genes in stable shLuc, shDDX3X#2, and shDDX3X#3 HepG2 cells. N = 3 per group. Representative images of DDX3X, DDB2, XPA, and Sp1 protein expression in shLuc-, shDDX3X#2-, and shDDX3X#3-knockdown HepG2 cells by Western blot analysis were shown (right plot). Protein levels were normalized to the loading control GAPDH, expressed as fold changes relative to controls (set as 1) and shown under their corresponding panels. C–E, Binding of DDX3X and Sp1 to the promoter regions of the DDB2 and XPA genes. ChIP experiments were performed with liver tissues from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice (C), DDX3X- (shLuc, shDDX3X#2, and shDDX3X#3) (D), and Sp1- (siControl and siSp1) knockdown HepG2 cells (E). The cross-linked protein–DNA complex immunoprecipitated with anti-DDX3X and anti-Sp1 antibodies from liver tissues, and cultured cells were processed for ChIP-qPCR assays. The binding ability on the promoter region was presented in the relative fold change to rabbit IgG, which was normalized with input individually. N ≥ 5 per group. Representative images of DDX3X, Sp1, DDB2, and XPA protein expression in siControl- and siSp1-knockdown HepG2 cells by Western blot analysis were shown (E, left plot). The relative levels of proteins were quantified and shown under their corresponding panels. *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

Figure 5.

The targets of DDX3X are involved in NER. A, qRT-PCR analyses of NER genes and Sp1 in livers from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice. N = 4 per group. B, Relative mRNA levels of DDX3X, DDB2, ERCC6, ERCC8, XPA, and Sp1 genes in stable shLuc, shDDX3X#2, and shDDX3X#3 HepG2 cells. N = 3 per group. Representative images of DDX3X, DDB2, XPA, and Sp1 protein expression in shLuc-, shDDX3X#2-, and shDDX3X#3-knockdown HepG2 cells by Western blot analysis were shown (right plot). Protein levels were normalized to the loading control GAPDH, expressed as fold changes relative to controls (set as 1) and shown under their corresponding panels. C–E, Binding of DDX3X and Sp1 to the promoter regions of the DDB2 and XPA genes. ChIP experiments were performed with liver tissues from 6-week-old controls and Alb-Cre/+;Ddx3xflox/flox mice (C), DDX3X- (shLuc, shDDX3X#2, and shDDX3X#3) (D), and Sp1- (siControl and siSp1) knockdown HepG2 cells (E). The cross-linked protein–DNA complex immunoprecipitated with anti-DDX3X and anti-Sp1 antibodies from liver tissues, and cultured cells were processed for ChIP-qPCR assays. The binding ability on the promoter region was presented in the relative fold change to rabbit IgG, which was normalized with input individually. N ≥ 5 per group. Representative images of DDX3X, Sp1, DDB2, and XPA protein expression in siControl- and siSp1-knockdown HepG2 cells by Western blot analysis were shown (E, left plot). The relative levels of proteins were quantified and shown under their corresponding panels. *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

Close modal

To gain mechanistic insight into the link between DDX3X and NER, the mRNA levels of the DDB2, ERCC6, ERCC8, and XPA genes in DDX3X-knockdown (shDDX3X#2 and shDDX3X#3) and control (shLuc) HepG2 cells (9) were investigated. Our results showed that DDX3X, DDB2, XPA, and Sp1, but not ERCC6 and ERCC8, were significantly decreased in DDX3X-knockdown cells (Fig. 5B). Accordingly, the DDX3X, DDB2, XPA, and Sp1 protein levels were significantly reduced in DDX3X-knockdown cells. Chromatin immunoprecipitation quantitative PCR (ChIP-qPCR) assays showed the recruitment of DDX3X proteins to the promoter regions of Ddb2 and Xpa was approximately 2.8-fold and 3.6-fold higher, respectively, in livers from 6-week-old controls than in Alb-Cre/+;Ddx3xflox/flox mutants. The recruitment of Sp1 proteins to the promoter regions of Ddb2 and Xpa showed decreasing trends in Alb-Cre/+;Ddx3xflox/flox livers compared with those of controls (Fig. 5C). Consistently, we detected significantly decreased recruitment of DDX3X proteins and slightly decreased recruitment of Sp1 proteins to the promoter regions of DDB2 and XPA in DDX3X-knockdown HepG2 cells (Fig. 5D). In addition, the recruitment of DDX3X and Sp1 proteins to the promoter regions of DDB2 and XPA was decreased in Sp1-knockdown (siSp1) HepG2 cells (Fig. 5E). These results confirmed that DDX3X can transcriptionally regulate the expression of NER factors DDB2 and XPA through cooperation with Sp1, thus suggesting an important role of DDX3X in NER.

Loss of Ddx3x results in genomic stress in rapidly dividing tumor cells

In normal cells, efficient DDR is crucial for the maintenance of the genome integrity. Previously, we showed inactivation of Ddx3x results in embryonic lethality due to widespread DNA damage and apoptosis (12). The data so far showed that inactivation of Ddx3x in juvenile hepatocytes results in defective DNA repair and sustained DNA damage, which are causally connected to tumor initiation. Given the rapid cell-cycle time is a common feature of affected Ddx3x-deficient embryonic cells and juvenile hepatocytes, we therefore proposed that the reduced NER activity in Alb-Cre/+;Ddx3xflox/flox hepatocytes may cause genomic stress in the tumor. Indeed, IHC analysis of liver sections from aged controls and tumor-bearing Alb-Cre/+;Ddx3xflox/flox mutants showed that cCasp3 (apoptosis)-, Ki67 (compensatory proliferation)-, and γH2AX-positive hepatocytes were significantly increased in tumors compared with adjacent nontumor liver tissues from Alb-Cre/+;Ddx3xflox/flox mutants and their controls. In addition, the numbers of cells positive for pATR and 53BP1 nuclear bodies were significantly increased in tumors from Alb-Cre/+;Ddx3xflox/flox mutants (Fig. 6). Altogether, our data suggest that Ddx3x plays a critical role in the maintenance of genome integrity in the cell populations with relatively rapid rates of proliferation.

Figure 6.

The percentages of cCasp3-, Ki67-, γH2AX-, 53BP1-, and pATR-positive hepatocytes are significantly increased in tumor samples from Alb-Cre/+;Ddx3xflox/flox mutants. The liver tissues from 18- to 24-month-old controls and tumor-bearing Alb-Cre/+;Ddx3xflox/flox mice were processed for immunostaining. A, Representative images of cCasp3 (red), Ki67 (red), γH2AX (green), 53BP1 (green), and pATR (green) immunostaining. The stained cells are shown at higher magnification in the inset. Nuclei were stained with hematoxylin and/or DAPI. B, Percentages of cCasp3-, Ki67-, γH2AX-, 53BP1-, and pATR-positive hepatocytes in liver sections were quantified. N = 3 per group. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bar, 200 μm.

Figure 6.

The percentages of cCasp3-, Ki67-, γH2AX-, 53BP1-, and pATR-positive hepatocytes are significantly increased in tumor samples from Alb-Cre/+;Ddx3xflox/flox mutants. The liver tissues from 18- to 24-month-old controls and tumor-bearing Alb-Cre/+;Ddx3xflox/flox mice were processed for immunostaining. A, Representative images of cCasp3 (red), Ki67 (red), γH2AX (green), 53BP1 (green), and pATR (green) immunostaining. The stained cells are shown at higher magnification in the inset. Nuclei were stained with hematoxylin and/or DAPI. B, Percentages of cCasp3-, Ki67-, γH2AX-, 53BP1-, and pATR-positive hepatocytes in liver sections were quantified. N = 3 per group. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bar, 200 μm.

Close modal

Loss of Ddx3x predisposes female mice to DEN-induced liver injury and tumorigenesis

The carcinogen, DEN, which is bioactivated by cytochrome P450 (Cyp) enzymes in the liver and generates DNA adducts, is widely used to induce liver injury and tumorigenesis in experimental animal models (29, 30). To examine the acute liver injury response, a single i.p. injection of DEN (100 mg/kg) was given to 10-week-old mice. We detected significantly increased ALT levels in Alb-Cre/+;Ddx3xflox/flox mice compared with those of littermate controls 1 day after DEN injection. At 3 days after DEN injection, the increased ALT levels in controls did not differ from those of the Alb-Cre/+;Ddx3xflox/flox mice (Supplementary Fig. S10A). Increased apoptosis (cCasp3) and compensatory proliferation (Ki67) matched the degree of liver injury in Alb-Cre/+;Ddx3xflox/flox mice when compared with controls 1 day after DEN injection. Notably, at both 1 and 3 days after DEN injection, γH2AX-positive hepatocytes were markedly increased in Alb-Cre/+;Ddx3xflox/flox mice compared with that of their littermate controls (Supplementary Fig. S10B and S10C). These results demonstrated that loss of Ddx3x facilitates DEN-induced DNA damage accumulation and liver injury.

We next injected a single dose of DEN (25 mg/kg) into Alb-Cre/+;Ddx3xflox/flox mice and their littermate controls at 2 weeks of age. We found that ALT levels were significantly elevated in the Alb-Cre/+;Ddx3xflox/flox mice, peaked at 1 month after injection, and then gradually declined at 3 and 6 months after injection, compared with the controls. ALT levels varied in the controls and Alb-Cre/+;Ddx3xflox/flox mice 9 months after injection (Fig. 7A). At 3 months after DEN injection, there was no perceptible tumor in livers from the Alb-Cre/+;Ddx3xflox/flox mice and their controls as assessed by macroscopic examination. Visible tumors were detected in 50.0% (6/12) of the Alb-Cre/+;Ddx3xflox/flox livers, whereas none (0/10) of their controls had tumors 6 months after injection. At 9 months after injection, 58.3% (7/12) of the Alb-Cre/+;Ddx3xflox/flox mice exhibited liver tumors, compared with an approximately 27.3 % (3/11) incidence of tumors in female controls (Fig. 7B). Alb-Cre/+;Ddx3xflox/flox mice had more macroscopically visible tumors than their controls. We noted the tumor sizes were not further increased in Alb-Cre/+;Ddx3xflox/flox livers, and the mean tumor sizes were not different between controls and Alb-Cre/+;Ddx3xflox/flox livers (Fig. 7C). Quantitative analyses of cCasp3-, Ki67-, and γH2AX-positive cells in immunostained liver sections further showed that all these indices in Alb-Cre/+;Ddx3xflox/flox mice 1 month after DEN injection were significantly higher than those of 6-week-old (age-matched) untreated Alb-Cre/+;Ddx3xflox/flox mice (Supplementary Fig. S11 compared with Fig. 3B–D). Notably, basophilic and alpha-fetoprotein (AFP)–positive nodules were readily detected in the Alb-Cre/+;Ddx3xflox/flox livers but not in their controls 3 months after DEN injection. The elevated expression of Afp, the AFP gene, in both nontumor and tumor tissues from Alb-Cre/+;Ddx3xflox/flox livers 6 months after DEN injection was further confirmed by qRT-PCR (Fig. 7D). Sirius red staining of the liver sections showed that DEN-treated Alb-Cre/+;Ddx3xflox/flox mice exhibited significant increases in fibrosis compared with controls 3 and 6 months after injection (Fig. 7E). In tumor-bearing controls and Alb-Cre/+;Ddx3xflox/flox mutants 9 months after DEN injection, IHC analysis of liver sections showed that the percentages of cCasp3-, Ki67-, γH2AX-, 53BP1-, and pATR-positive cells were significantly increased in tumors compared with adjacent nontumor liver tissues. We noted the percentages of γH2AX- and 53BP1-positive cells in tumor sections from DEN-treated controls and Alb-Cre/+;Ddx3xflox/flox mice were significantly higher than those of spontaneous tumors from aged Alb-Cre/+;Ddx3xflox/flox mice (Supplementary Fig. S12 compared with Fig. 6). All together, these results demonstrated that DEN-treated Alb-Cre/+;Ddx3xflox/flox mice had not only dramatically increased tumor multiplicity but also accelerated tumor progression with increased incidence of HCC. Therefore, a role for DDX3X as an important regulator of genome stability in vivo was confirmed.

Figure 7.

Loss of Ddx3x promotes DEN-induced liver tumorigenesis in Alb-Cre/+;Ddx3xflox/flox mice. Controls and Alb-Cre/+;Ddx3xflox/flox mice were i.p. injected with a single dose of DEN (25 mg/kg body weight) at 2 weeks of age. The sera and livers were collected at the indicated time intervals (months) after injection. A, The serum ALT levels. N ≥ 7 per group. B, Gross images of representative livers from DEN-treated mice. The numbers of tumor-bearing mice in each group are indicated in the bottom-left corners of the images. Tumors are indicated by arrows. The number of visible tumors per liver was counted. Scale bar, 1 cm. C, The average number and size of tumors in tumor-bearing controls and Alb-Cre/+;Ddx3xflox/flox mice. D, Representative H&E and AFP protein (red) staining of liver sections and quantitative mRNA levels of Afp in livers from control and Alb-Cre/+;Ddx3xflox/flox mutant 3 and 6 months after DEN injection. The boundaries of basophilic- and AFP-positive nodules, and tumors (T) were demarcated with dashed lines. Nuclei were counterstained with hematoxylin. Both nontumor (N) and tumor tissues were collected from Alb-Cre/+;Ddx3xflox/flox livers. N ≥ 5 per group. E, Representative sirius red staining of liver sections. Liver fibrosis was evaluated by sirius red staining. Quantification of the fibrotic area (4–8 fields, 100x magnification) per liver section was determined. N ≥ 6 per group. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bars, 1 cm (B); 200 μm (C and D).

Figure 7.

Loss of Ddx3x promotes DEN-induced liver tumorigenesis in Alb-Cre/+;Ddx3xflox/flox mice. Controls and Alb-Cre/+;Ddx3xflox/flox mice were i.p. injected with a single dose of DEN (25 mg/kg body weight) at 2 weeks of age. The sera and livers were collected at the indicated time intervals (months) after injection. A, The serum ALT levels. N ≥ 7 per group. B, Gross images of representative livers from DEN-treated mice. The numbers of tumor-bearing mice in each group are indicated in the bottom-left corners of the images. Tumors are indicated by arrows. The number of visible tumors per liver was counted. Scale bar, 1 cm. C, The average number and size of tumors in tumor-bearing controls and Alb-Cre/+;Ddx3xflox/flox mice. D, Representative H&E and AFP protein (red) staining of liver sections and quantitative mRNA levels of Afp in livers from control and Alb-Cre/+;Ddx3xflox/flox mutant 3 and 6 months after DEN injection. The boundaries of basophilic- and AFP-positive nodules, and tumors (T) were demarcated with dashed lines. Nuclei were counterstained with hematoxylin. Both nontumor (N) and tumor tissues were collected from Alb-Cre/+;Ddx3xflox/flox livers. N ≥ 5 per group. E, Representative sirius red staining of liver sections. Liver fibrosis was evaluated by sirius red staining. Quantification of the fibrotic area (4–8 fields, 100x magnification) per liver section was determined. N ≥ 6 per group. *, P < 0.05; **, P < 0.01; and ***, P < 0.001. Scale bars, 1 cm (B); 200 μm (C and D).

Close modal

Human HCC generally proceeds through a stepwise process involving hepatic injury and compensatory proliferation, followed by inflammation, fibrosis, and the development of HCC. Here, we showed that Alb-Cre/+;Ddx3xflox/flox mice recapitulated these key features of human HCC pathogenesis. Loss of Ddx3x affected DNA repair by reducing the expression of NER factors DDB2 and XPA, which contributed to an accumulation of unrepaired DNA damage and replication stress in the liver, providing further evidence that DDX3X maintains genomic stability in vivo.

In general, most adult tissues/organs are in a quiescent state, and proliferation occurs only when old/dying cells must be replaced. Nevertheless, the liver is a unique organ with a high regenerative capacity that can restore its lost mass after resections or injury. A cell-cycle kinetic study of the liver showed that the fraction of hepatocytes in a proliferative state, but not the average cell-cycle time, decreased gradually with age, beginning at around 2 weeks of age. As development proceeds, individual cells cease division and then increase in cell size until the final size limit is reached (31). In this study, we showed that the levels of RPA70, RPA32, pRPA32, and Mre11 were higher in livers from 3-week-old (juvenile) controls than those of 6-week-old controls, indicating that the expression of DDR factors during early liver growth is essential to safeguard DNA integrity. Indeed, mutations/inactivation of DDR genes are observed in various diseases and cancers (32). Among these studies, a higher incidence of spontaneous liver tumors and an increased mutation frequency in livers were observed in aged Xpa−/− mice compared with their wild-type controls (33, 34). In addition to Ultraviolet B (UVB)-induced skin carcinogenesis, Ddb2-deficient mice developed spontaneous tumors at a high rate between 20 and 25 months of age (35). Hepatocyte-specific ablation of Ddb1 leads to chronic hepatocyte turnover, mild liver damage, and liver tumor formation (36). In experimental models of alcohol-related liver diseases, the transglutaminase-2 (TG2)–mediated impairment of the Sp1-c-met signaling cascade was observed (37). These findings support our results that dysregulation of DDR, as evidenced by the decrease of Xpa and Ddb2 levels needed for DNA damage repair, in juvenile Alb-Cre/+;Ddx3xflox/flox livers is associated with a predisposition to HCC. This study shows the genomic instability is a driving event for liver tumorigenesis in Alb-Cre/+;Ddx3xflox/flox mutants; however, we do not exclude the possibility that DDX3X may also affect other cellular processes, such as epigenetic alterations, deregulation of miRNAs, and overexpression of oncogenes, during progression of HCC.

It is wildly accepted that inflammation can increase the risk of cancer by providing the cytokines and chemokines from infiltrating cells in tumor microenvironment, including HCC (38, 39). Targeting hepatic inflammation is one of the therapeutic opportunities for the treatment of HCC (40). In this study, we showed that the liver injury in 6-week-old hepatocytes and Alb-Cre/+;Ddx3xflox/flox mutants was associated with inflammation (Supplementary Fig. S4). The infiltrated immune cells and IL1β-, TNFα-, and NF-κB–positive cells were significantly increased in spontaneous tumor tissues of aged Alb-Cre/+;Ddx3xflox/flox mutants. Also, Il1β and Tnfα mRNA levels were significantly increased in liver tumors from Alb-Cre/+;Ddx3xflox/flox mutants. In tumor-promoting stimuli, IL1β- and TNFα-positive cells and Il1β levels were significantly higher in DEN-treated Alb-Cre/+;Ddx3xflox/flox tumors than those of controls (Supplementary Fig. S13). In the context of DNA damage and inflammation, DEN-treated Alb-Cre/+;Ddx3xflox/flox mice had not only dramatically increased tumor multiplicity but also accelerated tumor progression with increased incidence of HCC. These results suggested inflammation-related changes in the microenvironment of liver contribute to liver injury and tumorigenesis in Alb-Cre/+;Ddx3xflox/flox mice.

The epidemiologic data of human HCC show that males have a higher risk of developing liver tumors than females (3). However, we found that the tumor incidence in Alb-Cre/+;Ddx3xflox/flox female mice was higher than in Alb-Cre/+;Ddx3xflox/Y male mice in this study. DDX3X has a structural homolog, DDX3Y, located on Y chromosome (12, 41). Studies have shown that both of the DDX3X and DDX3Y genes are transcribed in multiple tissues. DDX3X protein was detected in all tissues analyzed; however, the DDX3Y protein was predominantly detected in male germ cells by translational control (42, 43). Expression and deletion analyses of DDX3Y suggested a specific function of DDX3Y in male fertility (44), whereas we have shown discrete and essential roles of DDX3X protein in mouse embryonic and placental development (12). These genetic data suggest that DDX3Y may act specifically in testis and may have a different function than DDX3X. Although Ddx3y mRNA levels in Alb-Cre/+;Ddx3xflox/Y livers were comparable with their controls at a young age, decreased Ddx3y levels were observed in aged Alb-Cre/+;Ddx3xflox/Y livers and tumors (Supplementary Fig. S14). We cannot rule out the possibility that DDX3Y acts as a functional substitute for the loss of DDX3X in some contexts. It will be of interest to assess whether basal Ddx3y mRNA and DDX3Y protein are involved in the gender differential effect of DDX3X in liver homeostasis and DEN-induced liver tumorigenesis.

HCC is a heterogeneous disease that shows high resistance to conventional chemotherapy and radiotherapy. It is generally believed that genomic instability may contribute to poor clinical outcomes (45, 46). This study suggested a link between DDX3X function and genome integrity in liver tumorigenesis, at least partly through the control of the expression of NER genes that mediate DNA repair. In the cell populations with relatively rapid rates of proliferation, including juvenile liver and tumor, loss of Ddx3x results in DNA damage, liver injury, cell death–compensatory proliferation, and inflammation. Under the condition of DNA damage and inflammation, aberration functions of DNA repair could promote the development of dysplastic lesion and subsequent HCC. Given that the small-molecule–mediated inhibition of DDX3X activity may provide new therapeutic opportunities to treat various viral infections and cancers (47, 48), the impact of DDX3X on DNA damage and repair pathways cannot be underscored. A detailed understanding of the biology, pathology, and the complexity of the cellular responses to DNA damage in HCC associated with DDR deficiency will allow further preclinical investigations and provide appropriate treatment strategies.

No potential conflicts of interest were disclosed.

Conception and design: C.-H. Chan, C.-M. Chen, L.-R. You

Development of methodology: C.-H. Chan, C.-M. Chen, L.-R. You

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C.-H. Chan, C.-M. Chen, L.-R. You

Writing, review, and/or revision of the manuscript: C.-H. Chan, C.-M. Chen, L.-R. You

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L.-R. You

Study supervision: C.-M. Chen, Y.-H. Wu Lee, L.-R. You

We thank Dr. Hao-Kang Li for technical support. The A6 BCM, TROMA-III (K19), and G8.8 (EpCAM) antibodies developed by V.M. Factor, R. Kemler, and A.G. Farr, respectively, were obtained from the Developmental Studies Hybridoma Bank, created by the NICHD of the NIH and maintained at The University of Iowa, Department of Biology, Iowa City, IA. We thank the Taiwan Animal Consortium (MOST107-2319-B-001-002)–Taiwan Mouse Clinic which is funded by the Ministry of Science and Technology (MOST) of Taiwan for technical support in collagen stain experiment. This work was supported by the Ministry of Science and Technology of Taiwan (grant numbers MOST103-2320-B-009-006-, MOST104-2320-B-009-001-, and MOST105-2320-B-009-001- to Y.-H. Wu Lee; MOST105-2320-B-010-002-, MOST106-2320-B-010-029-, and MOST107-2311-B-010-002- to L.-R. You); Center For Intelligent Drug Systems and Smart Bio-devices (IDS2B), National Chiao Tung University and Cancer Progression Research Center, National Yang-Ming University from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan; and the Ministry of Education in Taiwan, Aim for the Top University Plan to L.-R. You.

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.

1.
Shiraha
H
,
Yamamoto
K
,
Namba
M
. 
Human hepatocyte carcinogenesis (review).
Int J Oncol
2013
;
42
:
1133
38
.
2.
Kanda
M
,
Sugimoto
H
,
Kodera
Y
. 
Genetic and epigenetic aspects of initiation and progression of hepatocellular carcinoma.
World J Gastroenterol
2015
;
21
:
10584
97
.
3.
El-Serag
HB
. 
Epidemiology of viral hepatitis and hepatocellular carcinoma.
Gastroenterology
2012
;
142
:
1264
73
e1
.
4.
Karagozian
R
,
Derdak
Z
,
Baffy
G
. 
Obesity-associated mechanisms of hepatocarcinogenesis.
Metabolism
2014
;
63
:
607
17
.
5.
Aravalli
RN
,
Steer
CJ
,
Cressman
EN
. 
Molecular mechanisms of hepatocellular carcinoma.
Hepatology
2008
;
48
:
2047
63
.
6.
Schröder
M
. 
Human DEAD-box protein 3 has multiple functions in gene regulation and cell cycle control and is a prime target for viral manipulation.
Biochem pharmacol
2010
;
79
:
297
306
.
7.
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
2005
;
25
:
1991
2003
.
8.
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
.
9.
Li
HK
,
Mai
RT
,
Huang
HD
,
Chou
CH
,
Chang
YA
,
Chang
YW
, et al
DDX3 represses stemness by epigenetically modulating tumor-suppressive miRNAs in hepatocellular carcinoma.
Sci Rep
2016
;
6
:
28637
.
10.
Huang
JS
,
Chao
CC
,
Su
TL
,
Yeh
SH
,
Chen
DS
,
Chen
CT
, et al
Diverse cellular transformation capability of overexpressed genes in human hepatocellular carcinoma.
Biochem Biophys Res Commun
2004
;
315
:
950
58
.
11.
Su
CY
,
Lin
TC
,
Lin
YF
,
Chen
MH
,
Lee
CH
,
Wang
HY
, et al
DDX3 as a strongest prognosis marker and its downregulation promotes metastasis in colorectal cancer.
Oncotarget
2015
;
6
:
18602
12
.
12.
Chen
CY
,
Chan
CH
,
Chen
CM
,
Tsai
YS
,
Tsai
TY
,
Wu Lee
YH
, et al
Targeted inactivation of murine Ddx3x: essential roles of Ddx3x in placentation and embryogenesis.
Hum Mol Genet
2016
;
25
:
2905
22
.
13.
Postic
C
,
Shiota
M
,
Niswender
KD
,
Jetton
TL
,
Chen
Y
,
Moates
JM
, et al
Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase.
J Biol Chem
1999
;
274
:
305
15
.
14.
Postic
C
,
Magnuson
MA
. 
DNA excision in liver by an albumin‐Cre transgene occurs progressively with age.
Genesis
2000
;
26
:
149
50
.
15.
Weisend
CM
,
Kundert
JA
,
Suvorova
ES
,
Prigge
JR
,
Schmidt
EE
. 
Cre activity in fetal albCre mouse hepatocytes: utility for developmental studies.
Genesis
2009
;
47
:
789
92
.
16.
Schlageter
M
,
Terracciano
LM
,
D'Angelo
S
,
Sorrentino
P
. 
Histopathology of hepatocellular carcinoma.
World J Gastroenterol
2014
;
20
:
15955
64
.
17.
Williams
MJ
,
Clouston
AD
,
Forbes
SJ
. 
Links between hepatic fibrosis, ductular reaction, and progenitor cell expansion.
Gastroenterology
2014
;
146
:
349
56
.
18.
Shi
JH
,
Line
PD
. 
Effect of liver regeneration on malignant hepatic tumors.
World J Gastroenterol
2014
;
20
:
16167
77
.
19.
Mazouzi
A
,
Velimezi
G
,
Loizou
JI
. 
DNA replication stress: causes, resolution and disease.
Exp Cell Res
2014
;
329
:
85
93
.
20.
Zeman
MK
,
Cimprich
KA
. 
Causes and consequences of replication stress.
Nat Cell Biol
2014
;
16
:
2
9
.
21.
Marechal
A
,
Zou
L
. 
RPA-coated single-stranded DNA as a platform for post-translational modifications in the DNA damage response.
Cell Res
2015
;
25
:
9
23
.
22.
Toledo
LI
,
Altmeyer
M
,
Rask
MB
,
Lukas
C
,
Larsen
DH
,
Povlsen
LK
, et al
ATR prohibits replication catastrophe by preventing global exhaustion of RPA.
Cell
2013
;
155
:
1088
103
.
23.
Ciccia
A
,
Elledge
SJ
. 
The DNA damage response: making it safe to play with knives.
Mol Cell
2010
;
40
:
179
204
.
24.
Kamileri
I
,
Karakasilioti
I
,
Garinis
GA
. 
Nucleotide excision repair: new tricks with old bricks.
Trends Genet
2012
;
28
:
566
73
.
25.
Marteijn
JA
,
Lans
H
,
Vermeulen
W
,
Hoeijmakers
JH
. 
Understanding nucleotide excision repair and its roles in cancer and ageing.
Nat Rev Mol Cell Biol
2014
;
15
:
465
81
.
26.
Nichols
AF
,
Itoh
T
,
Zolezzi
F
,
Hutsell
S
,
Linn
S
. 
Basal transcriptional regulation of human damage-specific DNA-binding protein genes DDB1 and DDB2 by Sp1, E2F, N-myc and NF1 elements.
Nucleic Acids Res
2003
;
31
:
562
69
.
27.
Wang
Y
,
Li
F
,
Zhang
G
,
Kang
L
,
Guan
H
. 
Ultraviolet-B induces ERCC6 repression in lens epithelium cells of age-related nuclear cataract through coordinated DNA hypermethylation and histone deacetylation.
Clin Epigenetics
2016
;
8
:
62
.
28.
Topping
RS
,
Myrand
SP
,
Williams
BL
,
Albert
JC
,
States
JC
. 
Characterization of the human XPA promoter.
Gene
1995
;
166
:
341
42
.
29.
Heindryckx
F
,
Colle
I
,
Van Vlierberghe
H
. 
Experimental mouse models for hepatocellular carcinoma research.
Int J Exp Pathol
2009
;
90
:
367
86
.
30.
Umemura
A
,
Park
EJ
,
Taniguchi
K
,
Lee
JH
,
Shalapour
S
,
Valasek
MA
, et al
Liver damage, inflammation, and enhanced tumorigenesis after persistent mTORC1 inhibition.
Cell Metab
2014
;
20
:
133
44
.
31.
Chang
M
,
Parker
EA
,
Muller
TJ
,
Haenen
C
,
Mistry
M
,
Finkielstain
GP
, et al
Changes in cell-cycle kinetics responsible for limiting somatic growth in mice.
Pediatr Res
2008
;
64
:
240
45
.
32.
Jackson
SP
,
Bartek
J
. 
The DNA-damage response in human biology and disease.
Nature
2009
;
461
:
1071
78
.
33.
de Vries
A
,
van Oostrom
CT
,
Dortant
PM
,
Beems
RB
,
van Kreijl
CF
,
Capel
PJ
, et al
Spontaneous liver tumors and benzo[a]pyrene-induced lymphomas in XPA-deficient mice.
Mol Carcinog
1997
;
19
:
46
53
.
34.
Giese
H
,
Dolle
ME
,
Hezel
A
,
van Steeg
H
,
Vijg
J
. 
Accelerated accumulation of somatic mutations in mice deficient in the nucleotide excision repair gene XPA.
Oncogene
1999
;
18
:
1257
60
.
35.
Yoon
T
,
Chakrabortty
A
,
Franks
R
,
Valli
T
,
Kiyokawa
H
,
Raychaudhuri
P
. 
Tumor-prone phenotype of the DDB2-deficient mice.
Oncogene
2005
;
24
:
469
78
.
36.
Yamaji
S
,
Zhang
M
,
Zhang
J
,
Endo
Y
,
Bibikova
E
,
Goff
SP
, et al
Hepatocyte-specific deletion of DDB1 induces liver regeneration and tumorigenesis.
Proc Natl Acad Sci U S A
2010
;
107
:
22237
42
.
37.
Tatsukawa
H
,
Fukaya
Y
,
Frampton
G
,
Martinez-Fuentes
A
,
Suzuki
K
,
Kuo
TF
, et al
Role of transglutaminase 2 in liver injury via cross-linking and silencing of transcription factor Sp1.
Gastroenterology
2009
;
136
:
1783
95
e10
.
38.
Feng
GS.
Conflicting roles of molecules in hepatocarcinogenesis: paradigm or paradox.
Cancer Cell
2012
;
21
:
150
54
.
39.
Hagerling
C
,
Casbon
AJ
,
Werb
Z
. 
Balancing the innate immune system in tumor development.
Trends Cell Biol
2015
;
25
:
214
20
.
40.
Tahmasebi Birgani
M
,
Carloni
V
. 
Tumor microenvironment, a paradigm in hepatocellular carcinoma progression and therapy.
Int J Mol Sci
2017
;
18
:
405
.
41.
Yang
F
,
Babak
T
,
Shendure
J
,
Disteche
CM
. 
Global survey of escape from X inactivation by RNA-sequencing in mouse.
Genome Res
2010
;
20
:
614
22
.
42.
Ditton
HJ
,
Zimmer
J
,
Kamp
C
,
Rajpert-De Meyts
E
,
Vogt
PH
. 
The AZFa gene DBY (DDX3Y) is widely transcribed but the protein is limited to the male germ cells by translation control.
Hum Mol Genet
2004
;
13
:
2333
41
.
43.
Vong
QP
,
Li
Y
,
Lau
YF
,
Dym
M
,
Rennert
OM
,
Chan
WY
. 
Structural characterization and expression studies of Dby and its homologs in the mouse.
J Androl
2006
;
27
:
653
61
.
44.
Foresta
C
,
Ferlin
A
,
Moro
E
. 
Deletion and expression analysis of AZFa genes on the human Y chromosome revealed a major role for DBY in male infertility.
Hum Mol Genet
2000
;
9
:
1161
69
.
45.
Bouwman
P
,
Jonkers
J
. 
The effects of deregulated DNA damage signalling on cancer chemotherapy response and resistance.
Nat Rev Cancer
2012
;
12
:
587
98
.
46.
Singh
S
,
Singh
PP
,
Roberts
LR
,
Sanchez
W
. 
Chemopreventive strategies in hepatocellular carcinoma.
Nat Rev Gastroenterol Hepatol
2014
;
11
:
45
54
.
47.
Bol
GM
,
Xie
M
,
Raman
V
. 
DDX3, a potential target for cancer treatment.
Mol Cancer
2015
;
14
:
188
.
48.
Brai
A
,
Fazi
R
,
Tintori
C
,
Zamperini
C
,
Bugli
F
,
Sanguinetti
M
, et al
Human DDX3 protein is a valuable target to develop broad spectrum antiviral agents.
Proc Natl Acad Sci U S A
2016
;
113
:
5388
93
.

Supplementary data