Abstract
Hepatocellular carcinoma (HCC) is the fifth most lethal malignancy worldwide with no curative therapies. To discover potentially novel therapeutic targets for HCC, we previously studied the gene expression profiles of HCC patients and identified that significant upregulation of N-Myc downstream regulated gene 1 (NDRG1) is associated with more aggressive phenotypes and poorer overall survival of HCC patients. In this study, we further used a loss-of-function approach (RNA interference) to understand the role of NDRG1 in hepatocarcinogenesis. We found that suppression of NDRG1 significantly impaired HCC cell growth through inducing extensive cellular senescence of HCC cells both in vitro and in vivo, accompanied by cell cycle arrest at the G1 phase. The observed antitumor effects of NDRG1 suppression were correlated with activation of major senescence-associated signaling pathways, such as upregulation of tumor suppressors p53, p21 and p16, and decreased phosphorylated Rb. To obtain further insights into the clinical significance of NDRG1-modulated senescence in HCC patients, immunohistochemistry staining of 92 cases of HCC patients was done. We found that high NDRG1 expression (n = 66) is associated with low p21 (n = 82; P < 0.001) and low p16 (n = 86; P < 0.001) levels. In conclusion, this study demonstrated that NDRG1 is a potential therapeutic target for HCC because its suppression triggers senescence of HCC cells through activating glycogen synthase kinase-3β–p53 pathway, thereby inhibiting tumor progression.
Summary
We report a novel function of NDRG1 in regulating cellular senescence in HCC cells, via glycogen synthase kinase-3β–p53 pathway, suggesting that NDRG1 suppression is a likely therapeutic approach.
Introduction
Hepatocellular carcinoma (HCC), the primary form of adult liver cancer, is the fifth most lethal malignancy worldwide and the third leading cause of cancer-related deaths (1). The incidence and mortality rates of HCC are approximately equal, reflecting the dismal prognosis due to the difficulty in early diagnosis and lack of effective treatments (2). Treatment options are further limited by late presentation of the disease, due to its asymptomatic nature and the lack of sensitive and specific biomarkers for early detection. Additionally, the small percentage of patients who are suitable for surgical resection and the high postoperation recurrence rate lead to extremely low survival rate (3). To date, there is no effective chemotherapy for HCC. Thus, understanding the mechanisms underlying hepatocarcinogenesis may help to identify novel targets for more efficacious treatment strategies.
To identify novel therapeutic targets for HCC, we used cDNA microarrays to analyze the gene expression profiles in clinical HCC samples and uncovered a cluster of novel genes that are significantly upregulated or downregulated in HCC, including the N-Myc downstream regulated gene 1 (NDRG1) (4,5). Upregulation of NDRG1 transcript was found to be significantly associated with late-stage HCC, tumor invasion, poor survival and tumor recurrence (5,6). NDRG1 is a member of the NDRG family, which belongs to the α/β hydrolase superfamily (7,8), and it has been shown to play important roles in stress response, hormone response, cell growth and differentiation (9). NDRG1 expression can be stimulated by a variety of cellular stresses, especially hypoxia, which is a prevalent feature of solid tumors (10,11). In addition, NDRG1 is upregulated in mouse and human fetal liver tissue, but not expressed in the normal adult liver, suggesting its important role during embryonic liver development (12).
Overexpression of NDRG1 in tumor tissues and cancer cell lines has been reported in many cancers, including liver cancer (5,6), brain cancer (13), lung cancer (14) and gastric cancer (15). The tumor-specific upregulation of NDRG1 indicates that it is closely correlated with carcinogenesis. Furthermore, evidence that the hepatitis C virus nonstructural protein NS5A can interact with NDRG1 indicates its potentially important role in hepatocarcinogenesis (16). However, the underlying mechanism of how NDRG1 promote cancer remains elusive. In this study, we aimed to elucidate the contribution of NDRG1 overexpression toward hepatocarcinogenesis and to evaluate the potential of targeting NDRG1 as a prospective therapeutic approach for HCC.
Materials and methods
Cell lines and cell culture
HCC cell lines HepG2, Hep3B, Huh7, PLC/PRF/5, HepG2-Luc(+) and Hep3B-Luc(+) cells were cultured in Dulbecco’s Modified Eagle’s Medium with 10% fetal bovine serum (Invitrogen, Carlsbad, CA), and 1% penicillin-streptomycin, maintained in a humidified atmosphere of 5% CO2 at 37°C.
siRNA transfection and pifithrin-α treatment
Target-specific siRNA (Ambion, Austin, TX; 20nM each) was transfected into HCC cells using RNAimax transfection reagent (Invitrogen) according to manufacturer’s instructions. The protein knockdown efficiency was measured by western blotting for each target. The siRNA sequences are the following: NDRG1 siRNA-1: 5-GCU GAU CCA GUU UCC GGA Att-3; NDRG1 siRNA-2: 5-ACC UGC ACC UGU UCA UCA Att-3; p53 siRNA: 5-GAA AUU UGC GUG UGG AUG Att-3; p16 siRNA: 5-UGU CCU GCC UUU UAA CGU Att-3; glycogen synthase kinase-3β (GSK-3β) siRNA: 5-CGA GAG CUC CAG AUC AUG Att-3.
After transfected with NDRG1-siRNA, cells were treated with 5 µM of pifithrin-α (PFT-α, Sigma, St Louis, MO) for 48 h.
Cell viability assay
HCC cells were plated into 96 well plates at 4 × 103 cells/well for siRNA transfection. Cell viability was measured daily using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega, Madison, WI) according to the manufacturer’s instructions. Three independent experiments were done, each in triplicates.
Cell cycle analysis
Cells were usually collected 48 h posttransfection by trypsin digestion and centrifugation at 3000rpm. The cell pellets were resuspended in 70% ethanol at −20°C overnight. The cell cycle profiles were assessed by staining with 50 μg/ml propidium iodide (Sigma-Aldrich), 50 μg/ml RNase A (Sigma-Aldrich) and 0.1% Triton X-100 in phosphate-buffered saline. The intracellular propidium iodide fluorescence intensity of each population of 10 000 cells was measured in each sample using a BD LSR II FACS machine (BD Bioscience, San Jose, CA) at the Stanford FACS Core Facility.
Senescence-associated β-galactosidase and p53 immunofluorescence staining
Senescence β-Galactosidase Staining Kit (Cell Signaling Technology Inc., Danvers, MA) was used to assess senescence in cells. Cells (2 × 105) were plated in 6-well plates, then washed in phosphate-buffered saline and fixed for 10min at room temperature in 1% paraformaldehyde. Cells were then washed twice with phosphate-buffered saline and incubated at 37°C without CO2 with freshly prepared staining solution overnight. Five images (of about 50–100 cells each) of each group were taken by Nikon ECLIPSE 80I microscope (Nikon, Tokyo, Japan), and the number and percentages of cells that stained positive were calculated.
Cells were fixed with 4% paraformaldehyde in phosphate-buffered saline for 20min, permeabilized with 0.1% Triton X-100 for 15 min and stained for NDRG1 and p53 using specific antibodies (Cell Signaling Technology Inc.). Images of cells were taken by Nikon ECLIPSE 80I microscope (Nikon).
Protein extraction and western blotting
Tissues or cells (2 × 105 cells were plated in 6-well plates) were lysed with T-PER Tissue Protein Extraction Reagent (Thermo Fisher Scientific Inc., Rockford, IL) for 15min on ice and then centrifuged at 13 500rpm for 10min. Supernatants were collected for measurement of protein concentration. Protein lysates suspended in loading buffer were separated on 10% sodium dodecyl sulfate –polyacrylamide gels and then transferred onto nitrocellulose membranes for incubation with primary antibodies overnight at 4°C. This was followed by incubation with HRP-conjugated secondary antibodies. Immunocomplexes were detected by Supersignal West Pico Chemiluminescent Substrate (Thermo Fisher Scientific Inc.) according to the manufacture’s protocol. Primary antibodies: NDRG1, p53, p21, p-Rb, GSK-3β, GSK-3β 9ser, (Cell Signaling Technology Inc.), p16, glyceraldehyde 3-phosphate dehydrogenase (GAPDH, Santa Cruz Biotechnology, Santa Cruz, CA). Secondary antibodies: goat anti-rabbit and goat anti-mouse (Santa Cruz Biotechnology). Western blot images were quantified by Image J software, with GAPDH expression used as the internal control.
Lentiviral-based shRNA vector generation and cell transduction
NDRG1 siRNA-1 or negative control siRNA were purchased from Ambion, Inc. (Austin, TX). The EcoRI and XhoI fragment from the NDRG1-siRNA containing the hairpin structure was cloned into a doxycycline-inducible tet-on lentiviral vector. The correct clones were verified by sequencing. Lentiviral vectors were generated in HEK293T cells by using Trans-Lenti Packaging Kits (Thermo Fisher Scientific Inc.). Two MOI viral vectors were transduced into HepG2-Luc(+) and Hep3B-Luc(+) cells, and stable cells were selected using 3 μg/ml puromycin (Sigma-Aldrich). After addition of 1 μg/ml doxycycline (Sigma-Aldrich) for 48h, the NDRG1-shRNAmir or control-shRNAmir cells with equal TurboRFP expression of fluorescence intensity were sorted by FACS and used to generate xenografts in mice.
Subcutaneous xenograft in nude mice
HepG2-Luc(+) and Hep3B-Luc(+) cells (1 × 106 each) with control-shRNAmir or NDRG1-shRNAmir were resuspended in 100 µl matrigel (BD Biosciences) and injected into 4-week-old nude mice (Charles River Laboratories International Inc.,Wilmington, MA) to induce subcutaneous tumor formation. Cells stably expressing negative control-shRNAmir or NDRG1-shRNAmir were injected into the left shoulder or right shoulder, respectively (n = 8 for each group). One week postinjection, whole body luminescence imaging of nude mice was done to establish a baseline for tumor size. Doxycycline (200 μg/ml) was then added into the daily water of nude mice, and the tumor luminescence intensity and tumor diameters were measured weekly. The tumor volume in mm3 was calculated by the formula: volume = (width)2 × length/2. The mice were killed on week 4, and the tumor tissues were collected for further analysis.
Immunohistochemistry of patient samples
Human HCC were collected from HCC patients who underwent hepatectomy at Stanford Hospital (USA). The study was approved by the Institutional Review Boards at Stanford Hospital for the use of human subjects in research studies, and all patients signed informed consent forms for using their tissue specimens for research purposes. Clinical tissues were embedded in paraffin and sectioned at 5 μm thick for immunohistochemical staining. After sections were treated with hydrogen peroxide and blocking buffer (10% normal goat serum), primary antibodies were added to the sections and incubated at 4°C overnight. Thereafter, secondary antibodies were added to the sections and incubated at room temperature for 1 h. Sections were developed using Dako EnVision system and were then counterstained with hematoxylin (Dako, Glostrup, Denmark). Images were viewed with Nikon epifluorescent upright microscope E600 (Nikon). Slides were scored by three pathologists in a blind fashion, given a scoring system based on the signal intensities (1—negative; 2—low expression, positive cells present in < 50% of the entire area; 3—high expression, positive cells present in > 50% of the entire area).
Data analysis
Statistical analyses were performed using PRISM version 5.0 software (GraphPad Software, Inc., La Jolla, CA). The Student’s t-test, Fisher’s Exact test and one-way ANOVA were used for calculating the significance between different groups. Statistical significance is indicated by P < 0.05.
Results
Suppression of NDRG1 induces cellular senescence in HepG2 cells
To select proper in vitro and in vivo models to investigate the role of NDRG1 in hepatocarcinogenesis, we first measured the NDRG1 protein expression levels in a panel of human HCC cell lines by western blotting. Strong expression of NDRG1 was detected in the fast-growing HCC cells, HepG2 and Huh7, whereas the slower growing cells Hep3B and PLC/PRF/5 have relatively lower NDRG1 expression (Figure 1A, left panel; Supplementary Figures 1A and 2, available at Carcinogenesis Online). We then selected two NDRG1-specific siRNAs and examined the knockdown efficiency in HepG2 cells, which have the strongest NDRG1 expression. Both siRNAs completely abolished NDRG1 protein expression in HepG2 cells after 72h of transfection (Figure 1A, right panel; Supplementary Figure 1A, available at Carcinogenesis Online. P < 0.05). Viability of HepG2 cells was significantly reduced by about 50% after 96h transfection with NDRG1-specific siRNA compared with negative control siRNA group (Figure 1B, P < 0.05). Furthermore, HepG2 cells transfected with NDRG1-specific siRNAs showed significantly reduced percentages of S-phase cells, accompanied by an increase in the percentages of G1-phase cells (Figure 1C, P < 0.05).
Fig. 1.
Suppression of NDRG1 inhibits viability and induces senescence in HepG2 cells.
(A) NDRG1 expression was detected in a panel of HCC cell lines by western blotting (left panel). Transfection of HepG2 cells with NDRG1-specific siRNA-1 and siRNA-2 abolished NDRG1 protein expression (right panel). GAPDH was used as loading control. NDRG1-siRNA significantly (B) reduced cell viability of HepG2, determined by using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay; *P < 0.05; (C) caused cell cycle arrest at the G1 phase, with significantly reduced S-phase cells 48h posttransfection, determined by flow cytometry; *P < 0.05; (D) induced cellular senescence as detected by SA-β-Gal staining and increased the percentage of SA-β-Gal positive cells. Cells with blue perinuclear staining were defined as positive cells. The number of cells was counted in five randomly taken images of each group; *P < 0.05; (E) induced the expression of senescence-associated proteins p53, p21, p16, Rb and p-Rb, with GAPDH as loading control; and (F) induced nuclear accumulation of p53 by immunofluorescence staining. Nuclei were stained by 4',6-diamidino-2-phenylindole.
Based on the significantly impaired cell viability and morphological alterations, we hypothesized that NDRG1 suppression may induce cellular senescence in HCC cells. To confirm this, SA-β-Gal staining was used to detect senescence after NDRG1 suppression in HepG2 cells. A large percentage (about 80%) of HepG2 cells transfected with NDRG1-specific siRNA was positive for SA-β-Gal staining (Figure 1D). To further determine which senescence pathway was activated by NDRG1 suppression, we screened for the expression of senescence-associated proteins in HepG2 cells transfected with NDRG1-specific siRNA. Expression levels of p53, p21 and p16 were elevated in cells with NDRG1 suppression, whereas p-Rb level was decreased (Figure 1E, Supplementary Figure 1A, available at Carcinogenesis Online). Additionally, nuclear accumulation of p53 was detected upon NDRG1 suppression (Figure 1F). These data suggested that NDRG1 suppression could activate p53- and p16-mediated cellular senescence in HCC cells.
Cellular senescence induced by NDRG1 suppression is partially prevented by blocking GSK-3β–p53/p16 pathway in HepG2 cells
To identify key senescence proteins that may be involved in cellular senescence induced by NDRG1 suppression, we used siRNAs targeting candidate proteins, either alone or in combination, to determine whether cellular senescence induced by NDRG1 suppression could be rescued. Cell viability and the percentage of S-phase cells were significantly restored upon co-transfection of NDRG1-specific siRNA with siRNA against p53, p16, p53 and p16, or GSK-3β (Figure 2A and B). The percentage of SA-β-Gal positive cells was also reduced in HepG2 cells co-transfected with NDRG1-specific siRNA and siRNA against p53, p16, p53 and p16, or GSK-3β (Figure 2C and D). Western blot results indicated that co-transfection of NDRG1-specific siRNA and siRNA against p53, p16, or p53 and p16 increased p-Rb level, which was decreased by NDRG1-specific siRNA (Figure 2E; Supplementary Figure 1B, available at Carcinogenesis Online). Additionally, GSK-3β siRNA decreased GSK-3β, GSK-3β 9ser, p53, p21 and p16 levels, indicating that GSK-3β can stabilize p53 and p16 levels in HepG2 cells with NDRG1 suppression (Figure 2F; Supplementary Figure 1B, available at Carcinogenesis Online).
Fig. 2.
Reversal of NDRG1 knockdown induced cellular senescence through inhibiting GSK-3β–p53/p16 pathway in HepG2 cells.
(A) Proliferation assay indicated that cell viability was increased in HepG2 cells after co-transfection of NDRG1-siRNA and siRNA against p53, p16, p53 and p16, or GSK-3β. *P < 0.05. (B) Cell cycle arrest induced by NDRG1 suppression was partially prevented after co-transfection of NDRG1-siRNA and siRNA against p53, p16, p53 and p16, or GSK-3β, as determined by flow cytometry 48h posttransfection; *P < 0.05. (C) Cellular senescence induced by NDRG1 suppression was partially prevented after co-transfection of NDRG1-siRNA and siRNA against p53, p16, p53 and p16, or GSK-3β. Representative images are shown. (D) The percentages of SA-β-Gal positive cells are shown; *P < 0.05. (E and F) Western blot detection of senescence-associated proteins after co-transfection of NDRG1-siRNA and siRNA against p53, p16, p53 and p16, or GSK-3β.
Alternatively, we used the p53 transcription inhibitor PFT-α to determine whether senescence induced by NDRG1 suppression could be prevented. Consistently, a decrease in p21 (a p53 downstream target) and increased p-Rb were detected in cells co-treated with NDRG1-siRNA and PFT-α (Figure 3A; Supplementary Figure 1C, available at Carcinogenesis Online). Cell viability of NDRG1-suppressed cells was also significantly increased after co-treatment with PFT-α compared with control cells without PFT-α (Figure 3B; P < 0.05). Furthermore, addition of PFT-α also reduced the percentage of senescence cells induced by NDRG1 suppression (Figure 3C and D, P < 0.05). Concomitantly, the percentage of S-phase cells was significantly increased upon co-treatment of NDRG1-siRNA and PFT-α (Figure 3E; P < 0.05). These data indicated that the cellular senescence induced by NDRG1 suppression is mediated by the GSK-3β–p53/p16 axis.
Fig. 3.
Specific p53 inhibitor PFT-α partially prevented senescence induced by NDRG1 suppression.
(A) Western blot showing that PFT-α partially prevented the effects of NDRG1 suppression on the expression levels of p53, p21 and p-Rb. (B) Cell viability was significantly increased in cells co-treated with NDRG1-siRNA and PFT-α. (*P < 0.05). (C) The number of SA-β-Gal positive cells was decreased in HepG2 cells co-treated with NDRG1-siRNA and PFT-α. (D) The percentage of senescence cells was counted after PFT-α treatment (P < 0.05). (E) Cell cycle arrest induced by NDRG1 suppression was partially prevented after co-transfection of NDRG1-siRNA and PFT-α, determined by flow cytometry 48h posttransfection; *P < 0.05.
Suppression of NDRG1 in vivo inhibited tumor growth through inducing senescence of HCC cells
To further examine the effects of NDRG1 suppression on hepatocarcinogenesis in vivo, we first generated lentivirus-derived doxycycline tet-on control and NDRG1-shRNAmir expression vectors. HepG2-Luc(+) and Hep3B-Luc(+) cells were then transduced with these vectors, selected for stable NDRG1-shRNAmir expression and used to establish xenografts in nude mice. The animal experiment is outlined in Supplementary Figure 3A (available at Carcinogenesis Online). In vitro, cell lines that stably express NDRG1-shRNAmir showed drastically reduced expression of NDRG1 after doxycycline treatment (Supplementary Figure 3B, available at Carcinogenesis Online and Figure 1D). One week after tumor inoculation, luminescence imaging showed comparable size of tumors in both the control and NDRG1 knockdown groups (Supplementary Figure 3D, available at Carcinogenesis Online). Thereafter, mice were fed with water containing doxycycline, and tumor luminescence intensity and tumor size were measured every week. After 4 weeks, the tumor sizes of NDRG1-shRNAmir groups were significantly reduced compared with control groups (Figure 4A; P < 0.05). These data indicated that suppression of NDRG1 in HCC cells effectively reduced tumorigenicity in mice.
Fig. 4.
NDRG1 suppression activates senescence in HCC cells in vivo.
(A) Growth curves showing the effect of NDRG1 suppression in HepG2 and Hep3B xenograft formation. Each group included 8 mice. *P < 0.05. (B) Representative images showing SA-β-Gal staining of subcutaneous tumor tissues (top panel) and frozen sections (bottom panel). (C) Western blot detection of senescence-associated proteins p53, p21, p16, GSK-3β, GSK-3β 9ser and p-Rb in xenograft tissues. GAPDH was used as loading control. (D) Immunohistochemistry staining of paraffin sections of xenograft tissues showing that NDRG1 suppression is accompanied by increased p21 and p16.
At the end of the experiment, mice were killed, and tumor tissues were harvested. Staining of whole tumor tissues and tissue sections with SA-β-Gal showed much higher senescence levels in HepG2 NDRG1-shRNAmir and Hep3B NDRG1-shRNAmir derived xenografts (Figure 4B). Western blotting of total protein extracts consistently showed reduced expression of NDRG1 in these xenografts (Supplementary Figure 3C, available at Carcinogenesis Online and Figure 1D, P < 0.05), which was accompanied by decreased phosphorylated GSK-3β serine 9 and p-Rb, as well as increased p53, p21 and p16 levels (Figure 4C, Supplementary Figure 1C, available at Carcinogenesis Online). Immunohistochemistry confirmed that NDRG1 suppression was associated with increased p21 and p16 expressions in both HepG2 NDRG1-shRNAmir- and Hep3B NDRG1-shRNAmir-derived xenografts (Figure 4D).
Correlation of NDRG1 with p21 and p16 in HCC patient samples
To obtain further insights into the clinical significance of NDRG1-modulated senescence in HCC patients, we stained for NDRG1, p21 and p16 in 92 cases of HCC represented on tissue microarrays (5). Semiquantitative analysis of immunohistochemistry signal intensity confirmed that high NDRG1 expression (n = 66) is significantly correlated with low p21 (n = 82; P < 0.001) and low p16 (n = 86; P < 0.001) expressions (Table I), confirming the inverse correlation observed in vitro. Representative images are shown in Supplementary Figure 4, available at Carcinogenesis Online. The detailed expression profiles of NDRG1, p21 and p16 are listed in Supplementary Table 1, available at Carcinogenesis Online.
Table I.
Correlation of NDRG1 expression with p21 and p16 expressions in 92 HCC patients
| Scorea | NDRG1 | Number of patients, p21b | p16c |
|---|---|---|---|
| 1 and 2 | 26 | 82 | 86 |
| 3 | 66 | 10 | 6 |
aRepresentative images of signal intensities and score system are shown in Supplementary Figure 5B.
bComparison between NDRG1 and p21; P < 0.001.
cComparison between NDRG1 and p16; P < 0.001.
Discussion
NDRG1 is an attractive therapeutic target for HCC due to its specific upregulation in liver tumors and absence in normal liver tissues. In this study, we demonstrated that NDRG1 suppression decreased HCC cell viability, which was consistently associated with enhanced cellular senescence, which were in part mediated by the GSK-3β–p53/p16 pathway. These cellular and molecular changes induced by NDRG1 suppression led to overall growth inhibition, in vitro and in vivo. The significant inverse correlation observed in vitro between NDRG1 expression and p21 and p16 expressions was also observed in HCC patients.
NDRG1 has been shown to be a physiological substrate of GSK-3β because three copies of the repeat sequence (GTRSRSHTSE) located at the C-terminal of NDRG1 are targeted by GSK-3β (17). GSK-3β is a serine–threonine kinase involved in multiple cellular functions, such as energy metabolism and cell differentiation, which are associated with the stabilization of tumor suppressors p53 and p16 (18). Based on this knowledge, we hypothesized that GSK-3β may be involved in the regulation of senescence by NDRG1. Indeed, we observed that NDRG1 suppression decreased phosphorylated GSK-3β (inactive form) and phosphorylated Rb (inactive form) expression levels, accompanied by simultaneous increases of p53, p21 and p16 expressions in both cultured HCC cells and their xenografts. Our data suggest that activation of GSK-3β is an important mechanism underlying senescence induced by NDRG1 suppression in HCC cells. We propose that in normal liver cells lacking NDRG1 expression, GSK-3β maintains the expression levels of p53 and p16 by inhibiting their degradation, thereby maintaining the balance of life and senescence (Figure 5). However, in liver cancer cells, cellular stresses such as hypoxia and oncogene activation may stimulate NDRG1 expression, which inhibits GSK-3β activity and leads to p53 and p16 degradation, allowing the cell to escape cellular senescence stimulated by stress signals. These findings enhance our understanding of the molecular mechanisms of hepatocarcinogenesis and suggest that specific inhibition of NDRG1 expression or function may have therapeutic potential.
Fig. 5.
Schematic model illustrating the antisenescence effect of NDRG1 in HCC pathogenesis.
Most chemotherapeutic regimens rely on high doses of toxic compounds to completely destroy tumor cells and are frequently associated with severe side effects, drug resistance, recurrence and progression to advanced malignancy. Therapy-induced senescence, which drives cells irreversibly into a state of quiescence without inducing radical events such as cell death, has been reported to produce equivalent antitumor effects, with fewer severe side effects and increased tumor-specific immune activity (19). The induction of senescence by NDRG1 suppression in HCC cells suggests its potential as an alternative, more efficacious and safer therapeutic approach than conventional chemotherapy.
Like other types of cancers, HCC cells display the typical hallmarks of cancer such as enabling replicative immortality and sustaining proliferative signaling (20). To adapt to high levels of oncogenic signaling, HCC cells frequently disable their senescence- or apoptotic-inducing circuitry by mutating key genes such as TP53 and Rb (21). Therefore, reactivation of these disabled signaling pathways may have therapeutic effects (22,23). In HCC, the dysfunction of tumor suppressor p53 is a common phenomenon that leads to aggressive proliferation and drug resistance. p53 is mutated in more than 50% of HCC patients, with higher rates observed in Asia and Africa (24). Remarkably, we found that NDRG1 suppression could induce cellular senescence in both p53-wild-type and p53-null HCC cells (by increasing the expression of alternative tumor suppressors like p16 and p21). Therefore, therapeutic targeting of NDRG1 is not limited to HCC with wild-type p53.
Recent studies have implied that upregulation of NDRG1 is correlated with tumor progression in several types of cancers (5,6,13–15). For instance, high NDRG1 expression was significantly correlated with tumor angiogenesis and malignant progression together with poor prognosis in gastric cancer (15). Although upregulation of NDRG1 is documented in most cancers, a few studies have reported tumor suppressor functions of NDRG1. In prostate cancer, downregulation of NDRG1 was found to be associated with metastatic progression via activation of p21 expression (25). In addition, NDRG1 suppression in PC3 prostatic cancer cells failed to enter senescence (26), which is contrary to our observation that NDRG1 induced cellular senescence in HCC cells through increasing GSK-3β, p53, p21 and p16. The inverse correlation of NDRG1 with p21 and p16 was similarly observed in a cohort of HCC patients, implying that the antisenescence function of NDRG1 may be conserved in HCC patients. The expression pattern, level and function of NDRG1 are perhaps context dependent in different types of cancer cells. The possible mutation, phosphorylation and different isoforms of NDRG1 may account for these differences in its role in different types of cancers and cell lines.
In conclusion, our study demonstrates that inhibition of NDRG1 triggers the senescence pathway to inhibit HCC cell growth. Because NDRG1 expression is associated with aggressive HCC phenotypes, therapeutic targeting of NDRG1 (such as using nanoparticles to deliver NDRG1-specific siRNA) may be especially effective in late-stage patients with high NDRG1 expression.
Supplementary material
Supplementary Table 1 and Figures 1–5 can be found at http://carcin.oxfordjournals.org/
Funding
H. M. Lui Foundation; C. J. Huang Foundation; T. S. Kwok Liver Research Foundation.
Conflict of Interest Statement: None declared.
Supplementary Material
Glossary
Abbreviations:
- GAPDH
glyceraldehyde 3-phosphate dehydrogenase
- GSK-3β
glycogen synthase kinase-3β
- NDRG1
N-Myc downstream regulated gene 1
- PFT-α
pifithrin-α.
References
- 1. El-Serag H.B. (2011). Hepatocellular carcinoma. N. Engl. J. Med., 365, 1118–1127. [DOI] [PubMed] [Google Scholar]
- 2. Altekruse S.F., et al. (2009). Hepatocellular carcinoma incidence, mortality, and survival trends in the United States from 1975 to 2005. J. Clin. Oncol., 27, 1485–1491. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Majno P.E., et al. (2000). Primary liver resection and salvage transplantation or primary liver transplantation in patients with single, small hepatocellular carcinoma and preserved liver function: an outcome-oriented decision analysis. Hepatology, 31, 899–906. [DOI] [PubMed] [Google Scholar]
- 4. Chen X., et al. (2002). Gene expression patterns in human liver cancers. Mol. Biol. Cell, 13, 1929–1939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Chua M.S., et al. (2007). Overexpression of NDRG1 is an indicator of poor prognosis in hepatocellular carcinoma. Mod. Pathol., 20, 76–83. [DOI] [PubMed] [Google Scholar]
- 6. Cheng J., et al. (2011). NDRG1 as a biomarker for metastasis, recurrence and of poor prognosis in hepatocellular carcinoma. Cancer Lett., 310, 35–45. [DOI] [PubMed] [Google Scholar]
- 7. Qu X., et al. (2002). Characterization and expression of three novel differentiation-related genes belong to the human NDRG gene family. Mol. Cell. Biochem., 229, 35–44. [DOI] [PubMed] [Google Scholar]
- 8. Shaw E., et al. (2002). Identification of a novel class in the alpha/beta hydrolase fold superfamily: the N-myc differentiation-related proteins. Proteins, 47, 163–168. [DOI] [PubMed] [Google Scholar]
- 9. Melotte V., et al. (2010). The N-myc downstream regulated gene (NDRG) family: diverse functions, multiple applications. FASEB J., 24, 4153–4166. [DOI] [PubMed] [Google Scholar]
- 10. Salnikow K., et al. (2002). The regulation of hypoxic genes by calcium involves c-Jun/AP-1, which cooperates with hypoxia-inducible factor 1 in response to hypoxia. Mol. Cell. Biol., 22, 1734–1741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Sibold S., et al. (2007). Hypoxia increases cytoplasmic expression of NDRG1, but is insufficient for its membrane localization in human hepatocellular carcinoma. FEBS Lett., 581, 989–994. [DOI] [PubMed] [Google Scholar]
- 12. He J., et al. (2003). Correlation of NDRG1 gene with liver tissue differentiation and hepatocarcinogenesis. Beijing Da Xue Xue Bao., 35, 471–475. [PubMed] [Google Scholar]
- 13. Northcott P.A., et al. (2012). Subgroup-specific structural variation across 1,000 medulloblastoma genomes. Nature, 488, 49–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Wang D., et al. (2012). NDRG1/Cap43 overexpression in tumor tissues and serum from lung cancer patients. J. Cancer Res. Clin. Oncol., 138, 1813–1820. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Kawahara A., et al. (2011). Nuclear expression of N-myc downstream regulated gene 1/Ca(2+)-associated protein 43 is closely correlated with tumor angiogenesis and poor survival in patients with gastric cancer. Exp. Ther. Med., 2, 471–479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Ahn J., et al. (2004). Systematic identification of hepatocellular proteins interacting with NS5A of the hepatitis C virus. J. Biochem. Mol. Biol., 37, 741–748. [DOI] [PubMed] [Google Scholar]
- 17. Murray J.T., et al. (2004). Exploitation of KESTREL to identify NDRG family members as physiological substrates for SGK1 and GSK3. Biochem. J., 384(Pt 3), 477–488. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Box A.H., et al. (2004). Cell cycle kinase inhibitor expression and hypoxia-induced cell cycle arrest in human cancer cell lines. Carcinogenesis, 25, 2325–2335. [DOI] [PubMed] [Google Scholar]
- 19. Kang T.W., et al. (2011). Senescence surveillance of pre-malignant hepatocytes limits liver cancer development. Nature, 479, 547–551. [DOI] [PubMed] [Google Scholar]
- 20. Hanahan D., et al. (2011). Hallmarks of cancer: the next generation. Cell, 144, 646–674. [DOI] [PubMed] [Google Scholar]
- 21. Pineau P., et al. (2008). Chromosome instability in human hepatocellular carcinoma depends on p53 status and aflatoxin exposure. Mutat. Res., 653, 6–13. [DOI] [PubMed] [Google Scholar]
- 22. Lu W.J., et al. (2011). Induction of mutant p53-dependent apoptosis in human hepatocellular carcinoma by targeting stress protein mortalin. Int. J. Cancer, 129, 1806–1814. [DOI] [PubMed] [Google Scholar]
- 23. Lu W.J., et al. (2011). Mortalin-p53 interaction in cancer cells is stress dependent and constitutes a selective target for cancer therapy. Cell Death Differ., 18, 1046–1056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Hussain S.P., et al. (2007). TP53 mutations and hepatocellular carcinoma: insights into the etiology and pathogenesis of liver cancer. Oncogene, 26, 2166–2176. [DOI] [PubMed] [Google Scholar]
- 25. Kovacevic Z., et al. (2011). The metastasis suppressor, N-myc downstream regulated gene 1 (NDRG1), upregulates p21 via p53-independent mechanisms. Carcinogenesis, 32, 732–740. [DOI] [PubMed] [Google Scholar]
- 26. Kobayashi A., et al. (2011). Bone morphogenetic protein 7 in dormancy and metastasis of prostate cancer stem-like cells in bone. J. Exp. Med., 208, 2641–2655. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.