Ku80 Functions as a Tumor Suppressor in Hepatocellular Carcinoma by Inducing S-phase Arrest through a P53-dependent Pathway

Ku80 is a component of the protein complex called DNA-dependent protein kinase, which is involved in DNA double-strand break repair and multiple other functions. Previous studies revealed that Ku80 haplo-insufficient and poly (adenosine diphosphate-ribose) polymerase-null transgenic mice developed hepatocellular carcinoma (HCC) at a high frequency. The role of Ku80 has never been investigated in human HCC. Ku80 expressions in HCC and adjacent liver tissue were investigated by using immunohistochem-ical staining and western blot. Ku80 was transfected into a Ku80-deficient HCC cell line SMMC7721 cells, and the growth features of the Ku80-expressing cells and vector-transfected cells were studied both in vitro and in vivo. Cell cycle analysis and RNA interference were employed to investigate the mechanisms underlying the growth regulation associated with Ku80 expression. Ku80 was found frequently downregulated in HCC compared with adjacent liver tissue. Ku80 downregulation was significantly correlated with elevated hepatitis B virus-DNA load and severity of liver cirrhosis. Overexpression of Ku80 in SMMC7721 cells significantly suppressed cell proliferation in vitro and in vivo. Ku80 overexpression caused S-phase cell cycle arrest and was associated with upregulation of p53 and p21 CIP1/WAF1 , and the inhibition of p53 or p21 CIP1/WAF1 expression by RNA interference overcame the growth suppression and S-phase arrest in the Ku80-expressing cells. A novel mechanism was revealed that Ku80 functions as a tumor suppressor in HCC by inducing S-phase arrest through a p53-dependent pathway.


Introduction
Hepatocellular carcinoma (HCC) is the third most deadly cancer worldwide, with .500000 new cases emerging annually.Most HCC patients are diagnosed at advanced stages that preclude the optimal surgical treatment (1).For the 20-30% of patients with resectable tumors, the 5 years survival rates are only 30-40% (2).To develop potential effective treatments for HCC, recent efforts have focused on exploring the molecular mechanisms underlying the pathogenesis of HCC.HCC is closely linked to persistent infection with hepatitis B virus (HBV) or hepatitis C virus (HCV), the intake of a-flatoxin B-contaminated food and excessive alcohol consumption (3).Approximately 90-95% of HCC cases result from the biological consequences of persistent HBV and HCV infection (4).Many studies have indicated that HCC develops through a multistage process resulting from the accumulation of genetic changes (5), including p53, phosphatase and tensin homolog, Rb and E-and T-cadherin inactivation (6)(7)(8)(9)(10) as well as transforming growth factor-b, vascular endothelial growth factor and Ras activation (11)(12)(13).However, the detailed molecular mechanism for the pathogenesis of HCC is still largely unknown.
Ku80 is well known for its critical role in the repair of double-strand breaks (DSBs) (14).There are two pathways for DSB repair: homologous recombination and non-homologous end joining (NHEJ).NHEJ is the predominant mechanism of DSBs repair in higher eukaryotes, whereas single-cell organisms rely more heavily on homologous recombination.Ku80 was found to maintain genome stability by repairing DSBs through NHEJ (15).Misrepaired DSBs are the major DNA lesions that lead to chromosomal aberration, mutation or carcinogenesis.Apart from its important role in DNA repair, Ku80 has been implicated by many studies to be involved in other cellular processes, such as telomere maintenance, apoptosis and tumor suppression (16,17).Ku80 has been suggested to be a multifunctional caretaker gene that suppresses chromosomal aberrations and malignant transformation (18).Ku80 deletion increased the incidence of cancer in mice that carried mutations in additional genes, such as p53 (19).Approximately 20% of Ku80 À/À p53 þ/À mice developed a broad spectrum of cancers, including lymphoma, osteosarcoma and leiomyosarcoma by 40 weeks of age, and 100% of Ku80 À/À p53 À/À mice developed pro-B-cell lymphoma by 16 weeks of age.Nevertheless, studies have also indicated that overexpression of Ku80 may be associated with bladder cancer, cervical carcinoma, pancreatic cancer and gastric cancer (20)(21)(22).Once Ku80 was silenced by RNA interference (RNAi), the proliferation of cervical carcinoma and esophageal squamous carcinoma cells was inhibited (23,24).These studies suggested that Ku80 played diverse roles in a number of cancer types.Using a transgenic mouse model, the study of Tong et al. (25) indicated that Ku80 haplo-insufficiency (Ku80 þ/À ) in poly (adenosine diphosphate-ribose) polymerase-1 null (PARP-1 À/À ) mice promoted the development of hepatocellular adenoma and HCC without the need for any carcinogen administration.This study suggested that the deletion of DNA damage repair molecules, Ku80 and PARP-1, played important roles in the development of HCC in mice; however, PARP-1 was found to increase activation rather than inactivation in human HCC compared with adjacent liver tissue (26).The role of Ku80 in the pathogenesis and development of human HCC remains unknown.
In this study, Ku80 expression status was investigated in human HCC and adjacent liver tissue.The growth features of the Ku80-expressing clones and vector-transfected cells were compared both in vitro and in vivo.The potential mechanisms of growth regulation associated with Ku80 expression were further investigated.

Patients and specimens
One hundred pairs of human HCC and their corresponding adjacent liver samples were obtained from 2008 to 2010 from patients who underwent liver Abbreviations: DMEM, Dulbecco's modified Eagle's medium; DNA-PK, DNA-dependent protein kinase; DSB, double-strand break; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; NHEJ, nonhomologous end joining; PARP-1, poly (adenosine diphosphate-ribose) polymerase-1; PBS, phosphate-buffered saline; RNAi, RNA interference; siRNA, small interfering RNA.y These authors contributed equally to this work.
Ó The Author 2012.Published by Oxford University Press.All rights reserved.For Permissions, please email: journals.permissions@oup.comresection at the Hepatic Surgery Center of Tongji Hospital affiliated with Huazhong University of Science and Technology.HCC was confirmed pathologically, and all specimens were stored at À80°C until analysis.Informed consent had been obtained from all patients.Clinicopathologic characteristics for these patients, including age, sex, hepatitis history, alpha-fetoprotein, liver cirrhosis, tumor number, tumor size, differentiation, preoperative Child-pugh score, Union for International Cancer Control stage and vascular invasion were shown in Table I.Tumor differentiation was evaluated by the same pathologist according to the criteria proposed by Edmonson and Steiner (27).Liver cirrhosis was graded into three stages according to the criteria described previously by our group (28).HCC was staged according to the tumor-node-metastasis staging system of the Union for International Cancer Control.The medical ethics committee of Tongji Hospital approved this study.

Immunohistochemical staining
Antigen retrieval of the tissue sections was performed in boiling citrate buffer for 15 min.Peroxide blocking was conducted with 0.3% peroxide in absolute methanol.After the slides had been incubated with primary antibodies (1:100 dilution) at 4°C overnight and washed twice with phosphate-buffered saline (PBS), they were then incubated with secondary antibody (Dako, Denmark) at 37°C for 30 min.After washing, the color reaction was developed with diaminobenzidene work solution (Dako).Positive cells were identified as those with nuclear staining.The percentage of positive cells was calculated by dividing the number of positive cells by the total number of hepatocytes in least 10 randomly chosen non-overlapping high-power (Â400) fields for each case.Protein expression was graded on a scale from þ to þþþ.The grade of 'þ' was given to cases whose percentage of positive cells was 25%, including those with no positive cells.If the average percentage of positive cells was !25% but 50%, then the expression was graded as 'þþ', whereas the percentage of positive cells !50% was graded as 'þþþ'.

Western blot analysis
The cell cultures and tissues were lysed in RIPA lysis buffer (50 mM Tris-HCl at pH 8.0, 1% NP40, 0.1% sodium dodecyl sulfate, 0.5% sodium deoxycholate, 0.02% sodium azide and 150 mM NaCl) containing Protease Inhibitor cocktail (Roche, Switzerland) on ice.After the protein concentration was determined by a bicinchoninic acid Kit (Pierce), 60 lg proteins were separated on precasted 10% sodium dodecyl sulfate-polyacrylamide gels and then electrotransferred onto polyvinylidene difluoride membranes (Millipore) in transfer buffer.The blots were blocked in 5% non-fat milk and incubated overnight at 4°C with primary antibodies (anti-Ku80 and p53 antibodies at 1:500 dilution; other antibodies at 1:200).The blots were then incubated with horseradish peroxidase-conjugated secondary antibody at 1:5000 dilution for 1 h at 37°C.The signals were visualized using the enhanced chemiluminescence system (Pierce).Protein expression was quantified by densitometry and normalized to b-actin expression using Alpha View software.

Cell culture and transfection
The human HCC cell lines SMMC7721 with p53 wild-type background, Hep3B with p53-null background and PLC/PRF/5 with p53-mutant background were obtained from the Center for Type Culture Collection of China and maintained in Dulbecco's modified Eagle's medium (DMEM) (Gibco), supplemented with 10% fetal bovine serum (Hyclone) and antibiotics (Invitrogen).SMMC7721, Hep3B and PLC/PRF/5 cells were cultured in 12-well plates and transiently transfected with the Ku80 plasmid and empty vector plasmid, respectively, by using Lipofectamine 2000 (Invitrogen) following the manufacturer's protocol.At the indicated time points, the cells were harvested for use in the western blot analysis and cell proliferation assays.For stable transfections, 5 Â 10 5 cells per well were seeded in a six-well plate for 24 h.Plasmid DNA was delivered into the cells using Lipofectamine 2000 following the manufacturer's protocol.Briefly, 4 lg of plasmid DNA was mixed with 10 ll of Lipofectamine 2000 at room temperature for 30 min before seeding in a well containing 1 ml of Opti-MEM medium (Invitrogen).After culturing in medium containing 500 lg/ml of G418 (Calbiochem, Germany) for 3 weeks, the individual clones were isolated.The positive cell clones that stably expressed Ku80 protein were then determined using western blot.The clones stably expressing Ku80 protein were maintained in medium containing 250 lg/ml of G418 for further experiments.

Cell proliferation assay
To assess serum-induced cell growth in culture, 10 000 cells were plated overnight in triplicates in a 12-well plate, serum starved for 48 h and then stimulated with the complete or selected growth media containing 10% fetal bovine serum for additional 5 or 7 days.At each indicated time point, the cells were trypsinized and counted.The number of cells, based on the average count of the three wells, was compared among the different groups from a total of three independent experiments.

Soft agar colony formation assay
Soft agar colony formation assay was performed for the Ku80-expressing and the vector-transfected clones as well as the parental SMMC7721 cells as described previously (29).Briefly, 1000 cells were equally divided into four wells in a 24-well plate in medium containing 0.3% noble agar and grown for 14-21 days.The number of colonies was determined by direct counting using an inverted microscope (Nikon, Japan).

Flow cytometry for cell cycle and apoptosis analysis
For cell cycle analysis, 1 Â 10 5 Ku80-expressing or vector-transfected SMMC7721 cells were serum starved for 48 h after washing with PBS for three times.The cells were stimulated with DMEM containing 10% serum for 12, 24, 36 and 48 h, and the cells were harvested and washed with PBS.The cells were then fixed with 70% ethanol.Immediately prior to the analysis, the cells were incubated with fresh propidium iodide containing RNase A for 30 min at 37°C.A total of 10 4 cells were analyzed from each sample on a fluorescence-activated cell sorting Calibur flow cytometer (Becton Dickinson).For apoptosis analysis, 1 Â 10 5 cells per well were seeded in the six-well plates and incubated with DMEM containing 10% serum.After 48 h, floating and adherent cells were harvested and washed twice with pre-cold PBS.The cells were stained for 15 min with Annexin V-fluorescein isothiocyanate and propidium iodide in 500 ll binding buffer and then analyzed by flow cytometry within 1 h.Ku80 as tumor suppressor in HCC c-H2AX immunofluorescence and neutral comet assay c-H2AX immunofluorescence assay was performed as described previously (30).Briefly, after methanol fixation, cells were permeabilized, then blocked in 3% bovine serum albumin/PBS and incubated with c-H2AX monoclonal antibody at 1:500.Secondary antibodies labeled with Alexa 555 (Molecular Probes) were added at 1:1000.Coverslips were mounted onto slides using Hard Set Vectashield with 4#,6-diamidino-2-phenylindole (Vector Laboratories).Cells were processed for comet tail formation using neutral comet assay conditions according to the methods described previously (31).Comets were analyzed using the measurement tool in OpenLab software.

RNAi knockdown
Three 21-nucleotide small interfering RNA (siRNA) duplexes targeting different coding regions of human p53, p21 CIP1/WAF1 and their scrambled sequence siRNA (mock) were custom synthesized by Riobio Company (GuangZhou, China).Three siRNA duplexes targeting p53 were named sip53-1, sip53-2 and sip53-3, and the duplexes targeting p21 CIP1/WAF1 were named sip21-1, sip21-2 and sip21-3, respectively.For the RNAi knockdown, equal numbers of cells were seeded in the plates containing medium without antibiotics for 24 h prior to the transfection.The siRNAs were introduced into the cells using Lipofectamine 2000 in serum-free Opti-MEM, according to the manufacturer's instructions.The expression levels of p53 and p21 CIP1/WAF1 proteins were determined by western blot.The two most efficient siRNAs for knockdown were chosen for further experiments.The transfected cells were grown in complete medium at 37°C and 5% CO 2 and harvested at different time points for use in the proliferation assay and cell cycle analysis described previously.

Growth in athymic immunocompromised mice
The Ku80-expressing and vector-transfected clone cells as well as parental SMMC7721 cells were harvested and resuspended to 1 Â 10 7 cells/ml.We subcutaneously injected 1 Â 10 6 cells in a total volume of 100 ll into the right flank of 4-to 6-week-old male athymic nude mice.Eight mice were injected for each clone.Their tumors were monitored every 4 days by measuring the tumor size using a caliper.The tumor volume was calculated by the formula, V (cm 3 ) 5 L Â W 2 /2, where L represents the longest dimension and W the shortest dimension of the tumor (32).Forty-four days after injection, the mice were killed and their tumors removed and weighed.Tumor tissue fragments were fixed in 10% formalin.The expressions of Ku80, p53, p21, cyclin A, cyclin E and cdk2 in xenograft tumor tissues were detected by immunohistochemical staining as described previously.

Statistical analysis
All results represent the average from triplicate experiments, and all results are expressed as the mean ± standard derivation.The associations between categorical variables were assessed using the chi-square test or the Fisher's exact test.
Analysis of variance was performed to determine the statistical significance among the groups.And a value of P ,0.05 was considered statistically significant.

Results
Ku80 is frequently downregulated in HCC and the downregulation was significantly correlated with HBV infection and liver cirrhosis Ku80 expressions in the 100 cases of paired HCC tissues and their corresponding adjacent liver tissues were studied.Immunohistochemical staining indicated that Ku80 was expressed in the nucleus of hepatocytes in liver tissues, and loss or decreased expression of Ku80 was frequently observed in HCC tissues (Figure 1A).There was a significant difference in Ku80 expression between the HCC tissues and their corresponding adjacent liver tissues (Figure 1B; P , 0.01).
The western blot further confirmed that Ku80 protein expression in 61 HCC samples was lost or decreased compared with their adjacent liver tissues (Figure 1C).The correlation between Ku80 expressions and clinicopathologic parameters in the patients with HCC was shown in Table I.The Ku80 downregulation was significantly correlated with elevated serum HBV-DNA load (P 5 0.004) and the severity of liver cirrhosis (P 5 0.02).However, there was no significant correlation between Ku80 expressions and other clinicopathologic parameters, such as age, sex, tumor size, tumor multiplicity, differentiation,

Ku80 overexpression suppresses cell growth and colony formation in SMMC7721 cells
To study the effects of Ku80 expression on cell growth, the pcDNA3.1(þ)-myc-his-Ku80plasmid and the vector plasmid were transiently transfected into three HCC cell lines Hep3B, PLC/PRF/5 and SMMC7721, respectively.Cell proliferation of the Ku80-transfected cells, the vector-transfected cells and the parental cells was studied.Ku80-transfected SMMC7721 cells (a Ku80-deficient HCC cell line) grew at significantly slower rates than the vector-transfected control cells (P , 0.01).However, there was no significant difference in proliferation between the Ku80-tranfected cells and the vector-tranfected cells for both Hep3B and PLC/PRF/5 cell lines (P .0.05; Supplementary data are available at Carcinogenesis Online).Subsequently, SMMC7721 was chosen for further stable transfection study.Ku80-expressing and vector-transfected stable SMMC7721 cell clones were then generated.
The western blot analysis confirmed that Ku80 clones 18, 26 and 33 expressed high protein levels of Ku80, whereas the clones with the vector and the parental SMMC7721 cells lacked Ku80 expression (Figure 2A).The growth curves of the Ku80-expressing cells, the vector-transfected cells and the parental SMMC7721 cells were determined using the cell proliferation assay.As shown in Figure 2B, on days 5 and 7, Ku80-18 and Ku80-26 cells grew at significantly slower rates than the control cells (P , 0.01).The number of cells in the Ku80-expressing clones decreased by 62.0-71.4% on day 7 compared with those in the vector clone and SMMC7721 cells (Figure 2B).Furthermore, the soft agar assay suggested that the colony number of the Ku80-expressing clones was significantly lower by 45.5-54.5% than those of the vectortransfected clone and the parental cells (Figure 2C).The statistics graph showed that there was a significant difference among the different cell types (Figure 2D; P , 0.01).

Ku80 overexpression causes S-phase arrest in SMMC7721 cells
To explore the mechanism underlying the cell growth suppression caused by Ku80 overexpression, the cell cycle distributions of the Ku80-expressing and the vector-transfected SMMC7721 cells were analyzed.After synchronization through serum starvation for 48 h, the Ku80-18 clones and vector clones were stimulated with DMEM containing 10% serum for 0, 12, 24, 36 and 48 h.Flow cytometry analysis demonstrated significantly increased numbers of cells from the Ku80-18 clone in S phase at all time points (Figure 3A and B) compared with those in the vector clone (Figure 3A and C).Under serum-starved conditions, the average percentages of vector-transfected clone cells and Ku80-expressing clone cells in S phase were 26.73 ± 1.79 and 48.60 ± 3.89%, respectively.After serum stimulation for 12, 24, 36 and 48 h, the percentages of vector-transfected clone cells in S phase were 34.06 ± 2.23, 39.60 ± 5.18, 34.70 ± 4.26 and 32.72 ± 3.50%, respectively, whereas the corresponding percentages of Ku80expressing cells were 69.14 ± 5.88, 73.24 ± 6.02, 81.08 ± 7.89 and 76.87 ± 8.06%, respectively.There was a significant difference in cell percentages in S phase between the vector-transfected and the Ku80expressing cells (Figure 3D; P , 0.01).The expressions of cell cycle regulators associated with S-phase regulation were further investigated.The expression levels of cdk2 and cyclin A were significantly decreased in the Ku80-expressing cells (P , 0.05), whereas the expression levels of p53 and p21 CIP1/WAF1 significantly increased (P , 0.05).The expression level of cyclin E remained unchanged (P .0.05) (Figure 3E).
To investigate whether apoptosis is also involved in the Ku80-induced cell growth inhibition, flow cytometry was performed for apoptosis analysis.Our data indicated that the apoptotic rates in SMMC7721, the vector-transfected cells, Ku80-18 and Ku80-26 clone cells were 1.81 ± 0.15, 1.83 ± 0.25, 9.44 ± 1.52 and 9.26 ± 1.72%, respectively (Figure 3F).There was a significant difference in cell apoptotic rate between Ku80-18 or Ku80-26 clone cells and SMMC7721 or the vector-transfected cells (P , 0.01; Figure 3G).These data indicate that

Ku80 as tumor suppressor in HCC
apoptosis is also involved in the Ku80-induced cell growth inhibition but the apoptotic rates in the Ku80-expressing cells are very low ($10%).
Ku80 overexpression decreases DNA damage and increases the genomic stability of SMMC7721 cells c-H2AX assay indicated that the numbers of c-H2AX foci in the nuclei of Ku80-18 and Ku80-26 cells were significantly decreased compared with SMMC7721 cells or the vector-transfected cells (Figure 4A and B; P , 0.05).Neutral comet assay also indicated that Ku80-18 and Ku80-26 clones exerted a significantly lower number of comet tails indicative of DNA damage in comparison with the control SMMC7721 cells or the vector-transfected cells (Figure 4A and C; P , 0.05).Our data indicated that Ku80 overexpression was able to decrease DNA damage and increase the genomic stability of SMMC7721 cells.
The Ku80-induced growth suppression and cell cycle arrest depend on the p53 pathway Ku80 overexpression induced the upregulation of p53 and p21 CIP1/WAF1 in the SMMC7721 cells, suggesting that the Ku80-induced growth suppression and S-phase arrest might be associated with the p53 pathway.To verify this assumption, we employed the specific RNAi technique to suppress the expression of p53 and its downstream regulator p21 CIP1/WAF1 .After three siRNA duplexes specific for p53 or p21 CIP1/WAF1 were transfected into the Ku80-18 clones, the protein expression of p53 or p21 CIP1/WAF1 significantly decreased, but no change in p53 or p21 CIP1/WAF1 expression was observed in the scrambled siRNA (mock)-transfected Ku80-18 cells.The sip53-2, sip53-3, sip21-1 and sip21-3 siRNA duplexes exhibited higher suppressive efficiency and were subsequently chosen for further study (data not shown).The Ku80-18 clone transfected with different concentrations (30, 50 and 100 nmol/l) of sip53-2, sip53-3, sip21-1 or sip21-3 had decreased p53 or p21 CIP1/WAF1 expression, respectively, in a dose-dependent manner at 72 h after transfection (data not shown).Further study indicated that the sip53-2-and sip53-3-induced p53 suppression started 24 h after transfection and lasted for 168 h (Figure 5A), whereas the sip21-1-and sip21-3-induced p21 CIP1/WAF1 inhibition started 24 h after transfection and lasted for 120 h (Figure 5B).The cell cycle analysis indicated that S-phase arrest was overcome in the sip53-2-, sip53-3-, sip21-1-or sip21-3-transfected Ku80-18 cells because the percentages of these cells in S phase decreased to the same levels as those of the vector control cells after 72 h of culture.In addition, there was no significant difference between the percentages of mock and sip53-2-, sip53-3-, sip21-1-or sip21-3-transfected vector control cells in S phase at all time points after transfection (Figure 5C and D; P .0.05).Cell proliferation assay demonstrated that the proliferation of the Ku80-18 cells was significantly increased after transfection with sip53-2 or sip53-3 compared with the mock-transfected Ku80-18 cells or the vector-tranfected control cells on days 5 and 7 of culture (Figure 5E; P , 0.01).It also indicated that the cell proliferation of the sip21-1-and sip21-3-transfected Ku80-18 cells were increased compared with the mock-transfected Ku80-18 cells or vector-transfected control cells on days 4 and 5 of culture (Figure 5F; P , 0.01).There was no significant difference in cell numbers between the mock-and sip53-2-or sip53-3-transfected vector-transfected control cells as well as between the mock-and sip21-1-or sip21-3-transfected vector-transfected control cells at all time points after transfection (Figure 5E and F; P .0.05).

Ku80 overexpression suppresses xenografts tumor growth in nude mice
Thirty-two 4-to 6-week-old male athymic nu/nu mice were divided into four groups of eight mice each.The mice received 1 Â 10 6 cells of Ku80-expressing clones (Ku80-18 and Ku80-26), the vector-transfected clone and the parental SMMC7721 cells, respectively, by subcutaneous injection into the right flank.Tumor volumes were measured with calipers every 4 days after injection.As shown in Figure 6A, when the tumors were removed from the killed mice on day 44 after injection, the mean volumes of the tumors derived from the Ku80-18 (1.66 ± 0.39 cm 3 ) and Ku80-26 (1.50 ± 0.31 cm 3 ) clones were significantly smaller than those derived from the vector-transfected (3.19 ± 0.39 cm 3 ) or the parental SMMC7721 (3.04 ± 0.51 cm 3 ) cells (P , 0.01).In addition, compared with the control groups, Ku80-expressing subcutaneous tumors (Ku80-18 and Ku80-26) grew at significantly slower rates and were smaller at all time points examined since day 24 after injection (Figure 6B; P , 0.01).The weights of the tumors derived from the Ku80-18 and Ku80-26 cells were significantly lower than those from the vector-transfected clone and the parental cells when the mice were killed (Figure 6C; P , 0.01).Immunohistochemical staining study indicated that Ku80, p53 and p21 expressions were increased in the tumors derived from the Ku80-transfected cells compared with those in the tumors from the vector-transfected cells (P , 0.01), whereas cyclin A and cdk2 expressions were decreased (P , 0.05) and cyclin E expression in tissue remained unchanged (P .0.05) (Figure 6D).The alteration pattern of these protein expressions in xenograft tumor tissues was consistent with that of those protein expressions identified in vitro in the Ku80-expressing cells and the vector-transfected cells.

Discussion
Genome instability is the hallmark of all forms of cancers because the mammalian genome is at constant risk from genotoxic factors and accumulates various mutations until malignant transformation occurs (29).Chromosomal instability is one of the salient features in HCC development and has been observed in !90% of HCC patients (33).HCC is closely correlated with cirrhosis following infection with HBV or HCV, and liver cirrhosis is associated with genotoxic DNA damage and mutations of known DNA repair genes (34).Under chronic genotoxic stress, failure of liver cells to initiate DNA repair and growth control may contribute to liver carcinogenesis.Hepatitis B virus X protein, an important oncogenic molecule of HBV, has been shown to significantly inhibit the ability of cells to repair damaged DNA by downregulating a series of DNA repair molecules, such as Xeroderma pigmentosum group B, Xeroderma pigmentosum group D and human DNA glycosylase alpha (hMYHalpha) (35,36).HCV infection has also been implicated in impairing DNA damage repair by repair enzymes, including Ku70 and GADD45b (37,38).These studies suggested that infection with HBV or HCV may cause the downregulation of these DNA damage repair enzymes, and the disruption of the DNA damage repair process might be involved in the mechanism for HBV-or HCV-associated HCC carcinogenesis.DNA damage repair molecule Ku80 is a key molecule in repairing DSBs through NHEJ and maintaining genome stability, and previous studies revealed that the deletion of both DNA damage repair molecules, Ku80 and PARP-1, promoted HCC development in a transgenic mouse model (25).Decreased Ku80 has been reported to be involved in transforming growth factor-a/c-myc-associated hepatocarcinogenesis in a mouse model (39).Our study found that Ku80 expression was frequently downregulated in human HCC tissues compared with adjacent liver tissues, and Ku80 downregulation was significantly correlated with elevated HBV-DNA load and the severity of liver cirrhosis.Furthermore, the overexpression of Ku80 is one of the subunits of DNA-dependent protein kinase (DNA-PK).Previous studies showed that DNA-PK acts upstream of p53 in response to DNA damage, allowing p53 to bind to specific DNA sequences and preventing MDM2 from inhibiting the p53-dependent transactivation (40).DNA-PK downregulation by RNAi reduced the accumulation of p53 by affecting the stability of p53 through Akt/ Protein Kinase B and Glycogen Synthase Kinase-3 phosphorylation (41).A previous protein-protein interaction analysis suggested that the p53 DNA-binding domain may be an interacting partner of Ku80 in the nucleus (42).Our study indicated that the overexpression of Ku80 was associated with upregulation of p53 and p21 CIP1/WAF1 in SMMC7721 cells (with wild-type p53 background).In contrast, the study of Holcomb (43) indicated that Ku80 deletion led to p53 activation in the Ku80 À/À APC MIN mice and reduced multiple intestinal adenoma and adenocarcinoma development.Ku80 may be implicated to play a diverse role in a particular type of tissue or tumor.The underlying regulating mechanism between Ku80 and p53 still needs to be further defined in the future.The western blot showed that p53 suppression induced by sip53-2 and sip53-3 in the Ku80-18 clone started at 24 h after transfection and lasted for 168 h (A), whereas the p21 CIP1/WAF1 inhibition induced by sip21-1 and sip21-3 also started at 24 h after transfection and lasted for 120 h.(B) The cells transfected with scrambled siRNA served as controls (mock).The cell cycle analysis indicated that the percentages of vector cells transfected with sip53-2 and sip53-3 (C) or sip21-1 and sip21-3 (D) in S phase were comparable with that of the cells mock transfected at all time points.In contrast, the percentages of Ku80-18 cells in S phase significantly decreased past 48 h after transfection with sip53-2 and sip53-3 or sip21-1 and sip21-3 (P , 0.01).(E and F) The cell proliferation assay further showed that the knockdown of p53 or p21 CIP1/WAF1 by sip53-2 and sip53-3 or sip21-1 and sip21-3 overcame the growth suppression induced by Ku80 in the Ku80-18 clone cells after days 5 or 4 of culture (P , 0.01).

S.Wei et al.
In addition, when p53 or p21 CIP1/WAF1 expression was inhibited by using RNAi, the S-phase arrest and cell growth suppression induced by the overexpression of Ku80 were overcome.On the contrary, Ku80 overexpression was unable to suppress cell proliferation in p53-null (Hep3B cell line) or p53-mutant (PLC/PRF/5 cell line) cells.These observations strongly suggested that the S-phase arrest and cell growth suppression induced by Ku80 overexpression in SMMC7721 cells were dependent on the p53 pathway.These results are supported by a previous study that suggested that DNA-PK was involved in the downregulation of cell proliferation and cell cycle progression (G 0 -to-S transition) through E2F-1-responsible genes (44).DNAdependent protein kinase catalytic subunit, a subunit of DNA-PK, has also been demonstrated to function as negative feedback to prevent excessive growth and tumor formation through the negative regulation of Akt upon fibroblast growth factor-2 treatment (45).Moreover, the p53 tumor suppressor was reported to inhibit cellular proliferation by inducing cell cycle arrest and apoptosis in response to cellular stresses (46).Our study has indicated that apoptosis is also involved in the Ku80-induced tumor growth inhibition.Considering the low apoptotic rates in the Ku80-expressing cells, cell apoptosis would only play a minor role in the Ku80-induced cell growth inhibition.Cell cycle arrest plays a major role in the Ku80-induced cell growth suppression in SMMC7721 cells.
Our data showed that Ku80 downregulation was significantly correlated with elevated serum HBV-DNA load and the severity of liver cirrhosis.HBV is the most common cause of cirrhosis in most of the countries (47).It was believed that cirrhosis is a consequence of the immune-mediated liver damage induced by chronic HBV infection (48).About 10-30% of patients with chronic hepatitis B developed liver cirrhosis in the end-stage of disease and at least 90% of HCCs are associated with the liver cirrhosis (47,49).However, the molecular mechanism for the Ku80 downregulation in human HCC is still not

Fig. 1 .
Fig. 1.Ku80 expression is frequently downregulated in HCC.(A) Representative immunostaining of Ku80 expression in human HCC tissues and corresponding adjacent liver tissues.Arrows indicate positive nuclear staining for Ku80.(B) Semiquantitative analysis of Ku80 expressions in the 100 cases of paired HCC tissues and their corresponding adjacent liver tissues.The P-value , 0.01 corresponds to the comparison of Ku80 expression between the HCC tissues and corresponding adjacent liver tissues.(C) Western blot showing Ku80 expression in HCC tissues (T) and corresponding adjacent liver tissues (N) from HCC patients, and b-actin was employed as an internal control.

Fig. 2 .
Fig. 2. Overexpression of Ku80 in SMMC7721 cells significantly inhibits cell growth in vitro and colony formation in soft agar.(A) Western blot showing Ku80 expression in the Ku80-transfected SMMC7721 clones, Ku80-18, Ku80-26 and Ku80-33, and no Ku80 expression in the vector-transfected clone or parental SMMC7721 cells.b-actin was included as a loading control for each sample.(B) Direct cell counting indicated significant suppression of proliferation of the Ku80-expressing clones Ku80-18 and Ku80-26 on days 5 and 7 of culture compared with that of the vector-transfected clone and parental SMMC7721 cells (P , 0.01).The mean and standard deviation are shown for each cell line.(C) Photomicrographs of representative fields showing that the Ku80-expressing SMMC7721 cell clones (Ku80-18 and Ku80-26) formed fewer and smaller colonies in soft agar than those of the vector-transfected clone and parental SMMC7721 cells.(D) Quantification of colony number.The asterisk denotes a statistically significant decrease in colony number ( Ã P , 0.01).The results shown represent the average value and standard deviation of triplicate experiments.

Fig. 3 .
Fig. 3. Ku80 overexpression results in a consistent increase in the percentages of cells in S phase.(A) The Ku80-expressing clone (Ku80-18) and the vectortransfected clones (vector) were synchronized by serum starvation for 48 h and then stimulated with DMEM containing 10% serum for 0, 12, 24, 36 and 48 h.The distribution of cells in G 1 , S and G 2 /M phases is represented graphically for Ku80-18 (B) and vector cells (C) after serum starvation and at each time point after serum stimulation.(D) At each time point after serum stimulation, significantly more Ku80-18 cells were in S phase than vector control cells (P , 0.01), even after serum starvation (P , 0.05).The result represents the average value and standard deviation from triplicate experiments.(E) Western blot showing the expression levels of Ku80, p53, p21, cyclin A, cyclin E and cdk2 in Ku80-expressing cells and control cells, whereas b-actin was included as a loading control for each sample.(F) Flow cytometry study indicating the apoptotic rates in Ku80-expressing cells and control cells.(G) There was a significant difference in cell apoptotic rate between Ku80-18 or Ku80-26 clone cells and SMMC7721 or the vector-transfected cells ( Ã P , 0.01).

Fig. 5 .
Fig.5.Ku80-induced growth suppression and cell cycle arrest depends on the p53 pathway.The western blot showed that p53 suppression induced by sip53-2 and sip53-3 in the Ku80-18 clone started at 24 h after transfection and lasted for 168 h (A), whereas the p21 CIP1/WAF1 inhibition induced by sip21-1 and sip21-3 also started at 24 h after transfection and lasted for 120 h.(B) The cells transfected with scrambled siRNA served as controls (mock).The cell cycle analysis indicated that the percentages of vector cells transfected with sip53-2 and sip53-3 (C) or sip21-1 and sip21-3 (D) in S phase were comparable with that of the cells mock transfected at all time points.In contrast, the percentages of Ku80-18 cells in S phase significantly decreased past 48 h after transfection with sip53-2 and sip53-3 or sip21-1 and sip21-3 (P , 0.01).(E and F) The cell proliferation assay further showed that the knockdown of p53 or p21 CIP1/WAF1 by sip53-2 and sip53-3 or sip21-1 and sip21-3 overcame the growth suppression induced by Ku80 in the Ku80-18 clone cells after days 5 or 4 of culture (P , 0.01).

Fig. 6 .
Fig. 6.Ku80 overexpression inhibits SMMC7721 cell-derived tumor growth in vivo.(A) The subcutaneous tumors derived from the Ku80-18 and Ku80-26 clones were smaller in size than those from the vector-transfected clones and parental SMMC7721 cells.(B) The tumor growth curve showed that the tumors derived from the Ku80-18 and Ku80-26 cells grew significantly slower than those from the control cells at all time points past 24 days after injection (P , 0.01).The mean and SD for tumor volumes were determined for each group.(C) When the mice were killed at 44 days after injection, the tumor weights of the Ku80-18 and Ku80-26 groups were lower than those of the control groups ( Ã P , 0.01).(D) The expression status of Ku80, p53, p21, cyclin A, cyclin E and cdk2 in the xenograft tumor tissues.Arrows indicate positive nuclear staining for these molecules.

Table I .
The correlation between Ku80 downregulation and clinicopathologic parameters in the patients with HCC AFP, alpha-fetoprotein; UICC, Union for International Cancer Control.a Chi-square test or the Fisher exact test.b Statistically significant (P , 0.05).