Negative regulation of transcription factor FoxM1 by p53 enhances oxaliplatin-induced senescence in hepatocellular carcinoma

Cancer Letters 331 (2013) 105–114

Contents lists available at SciVerse ScienceDirect

Cancer Letters journal homepage: www.elsevier.com/locate/canlet

Negative regulation of transcription factor FoxM1 by p53 enhances oxaliplatin-induced senescence in hepatocellular carcinoma Kai Qu a,1, Xinsen Xu a,1, Chang Liu a,⇑, Qifei Wu b, Jichao Wei a, Fandi Meng a, Lei Zhou a,c, Zhixin Wang a, Lei Lei d, Peijun Liu e a

Department of Hepatobiliary Surgery, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, 277 West Yanta Road, Xi’an 710061, China Department of Thoracic Surgery, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, 277 West Yanta Road, Xi’an 710061, China Department of General Surgery, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, 277 West Yanta Road, Xi’an 710061, China d Department of General Surgery, Gaoxin Hospital, Xi’an 710061, China e Translational Medical Center, First Affiliated Hospital of Medical College, Xi’an Jiaotong University, 277 West Yanta Road, Xi’an 710061, China b c

a r t i c l e

i n f o

Article history: Received 19 September 2012 Received in revised form 20 November 2012 Accepted 10 December 2012

Keywords: Forkhead box M1 p53 Oxaliplatin Cellular senescence Hepatocellular carcinoma

a b s t r a c t Previous studies have demonstrated the involvement of transcriptional factor forkhead box M1 (FoxM1) in cellular senescence of hepatocellular carcinoma (HCC). In the present study, we revealed that oxaliplatin could induce senescence in HCC cells, since advanced HCC patients with lower expression of FoxM1 were more sensitive to oxaliplatin therapy. Our data indicated that due to the repression by p53, FoxM1 played a critical role in oxaliplatin-induced senescence via regulating cycle-related proteins p21, p27, cyclins B1 and D1. Furthermore, inhibition of FoxM1, combined with oxaliplatin treatment, could significantly promote the senescence of HCC cells. Taken together, our findings suggest that FoxM1 may represent a promising therapeutic target for the medication of the chemosensitivity to oxaliplatin in HCC patients. Ó 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Hepatocellular carcinoma (HCC), the fifth most common cancer worldwide, is rarely detected early and is usually fatal within months of diagnosis, resulting in 500,000 deaths per year [1]. Despite implementation of potentially curative treatments such as liver resection and orthotopic liver transplantation, the prognosis is generally poor, as only 10–37% of patients are suitable surgical candidates because of advanced tumor or poor hepatic functional reserve [2]. Thus, chemotherapy is chosen to be the first-line treatment for advanced patients, despite the fact that HCC is a relatively resistant tumor with response rates ranging from 0% to 29% in randomized controlled trials [3]. Oxaliplatin, a third-generation platinum-derived chemotherapy agent, displays a wide spectrum of in vitro cytotoxic and in vivo antitumor activities. As an alkylating agent that causes DNA damage, oxaliplatin has been shown in clinical studies to have activity against advanced or metastatic HCC [4]. Although the precise mechanism of action is unknown, platinum compounds in general are thought to exert their therapeutic effects via induction of ⇑ Corresponding author. Tel./fax: +86 29 82653905. 1

E-mail address: [email protected] (C. Liu). These authors contributed equally to this work.

0304-3835/$ - see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.canlet.2012.12.008

various signal-transduction pathways, including DNA-damage recognition and repair, cell-cycle arrest, and apoptosis [5]. Recently, researchers have demonstrated that cancer cells derived from solid tumors would readily undergo senescence when exposed to a wide variety of DNA-damaging drugs, and among which, platinum compounds have also been shown to promote accelerated senescence [6–8]. Although oxaliplatin has been proven to be a promising systemic and locoregional chemotherapeutic agent for treatment of advanced HCC, patient response to oxaliplatin varies greatly depending on different conditions and only a small fraction of patients benefit greatly from this antitumor drug. Interestingly, in our previous studies, we accidently found that patients with lower expression level of transcriptional factor forkhead box M1 (FoxM1), which is a proliferation-associated transcription factor that has important roles in cellular proliferation, organogenesis, aging and tumorigenesis, responded well to oxaliplatin therapy, while those with higher expression level of FoxM1 did not. FoxM1, a member of the forkhead box transcription factor family, is characterized by the forkhead box domain. Previous studies have demonstrated that the overexpression of FoxM1 is associated with the development and progression of various cancers, such as lung, prostate, and pancreatic cancers as well as glioblastomas [9–12]. Recent studies have also suggested that FoxM1 plays a

106

K. Qu et al. / Cancer Letters 331 (2013) 105–114

crucial role in oxidative stress-induced cellular senescence [13– 15]. However, the role of FoxM1 in chemotherapy-induced senescence is still unknown. In the present study, we employed HepG2 and SMMC-7721 cells to investigate the possible mechanism underlying the FoxM1-mediated response to oxaliplatin treatment in hepatocellular carcinoma patients. We found p53-depended senescence was a potential therapeutic effect of oxaliplatin and that high FoxM1 expression would counter this therapeutic effect.

2.6. Senescence b-galactosidase (SA b-gal) assay

2. Materials and methods

2.7. RNA isolation and quantitative reverse transcription–polymerase chain reaction (qRT–PCR) analysis

2.1. Patients, samples and follow-up A total of 91 HCC patients who underwent oxaliplatin treatment at the First Affiliated Hospital of Xi’an Jiaotong University (Xi’an, China) were enrolled in this study. Forty-nine of the patients with advanced HCC underwent systemic oxaliplatin chemotherapy. Standard oxaliplatin treatment procedure was administered until unacceptable toxicity, disease progression, or intercurrent illness [4]. The other 42 HCC patients received preoperative hepatic arterial infusion (HAI) with oxaliplatin for reducing tumor size. When the suppressed tumor was within surgical indications, hepatectomy was performed. In addition, another 25 HCC patients who received hepatectomy without chemotherapy history were collected as control. HCC samples obtained from hepatectomy were mainly examined for senescence in liver tissues. The clinical characteristics of the patients were listed in Table 1. Ethical approval was obtained from the research ethics committee of First Affiliated Hospital, and written informed consent was obtained from each patient. The follow-up data were summarized at the end of December 2011 with a median follow-up of 11.5 months (range 2–45 months). In this study, the HCC patients who underwent HAI with oxaliplatin were excluded from survival analysis due to the short period of follow-up. Time to recurrence (TTR) and overall survival (OS) were considered as the primary endpoints. TTR was calculated as the time from registration to the first observation of disease recurrence, whereas OS was from registration to death or the final follow-up. 2.2. Cell culture and transfection Human HCC cell lines (Hep3B, HepG2, MHCC-97H and MHCC-97L) and immortal normal hepatocytes (L02) were obtained from Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences (Shanghai, China). SMMC-7721 cells were provided by the Molecular Biology Center of the First Affiliated Hospital, Xi’an Jiaotong University. FoxM1 siRNA (50 -GGACCACUUUCCCUACUUU-30 ), p53 siRNA (50 -CUGGAAGACUCCAGUGGUA-30 ) and p21 siRNA (50 -CGUCAGAACCCAUGCGGCA-30 ) were synthesized by Shanghai Gene Pharma Co. (Shanghai, China). FoxM1 overexpression plasmid (pcDNA3.1-FoxM1) was donored by Dr. Qichao Huang from the Fourth Military Medical University (Xi’an, China). Transfection of siRNA (100 nM) or plasmid (4 lg) was carried out using a Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the procedure recommended by the manufacturer. 2.3. MTT assay for cell viability HepG2 and SMMC-7721 cells (5.0  104 cells/well) were seeded in 96-well plates and incubated with oxaliplatin at various concentrations (0, 1.25, 2.5 and 5.0 lg/mL) for 24, 48 or 72 h at 37 °C. Then, 20 lL of MTT solution (5 g/L) was added into each well and incubated for another 4 h. Supernatants were removed and formazan crystals were dissolved in 200 mL dimethylsulfoxide (DMSO). Finally, optical density was determined at 490 nm using a POLAR star OPTIMA microplate reader (BMG Labtechnologies, Ortenberg, Germany). 2.4. Colony-forming assay Cellular proliferation was determined using a colony-forming assay. Twentyfour hours after oxaliplatin treatment, HepG2 cells were plated in 6-well tissue culture plates with a cell density of 5  103 cells/well. After 2 weeks of incubation, colonies were stained with crystal violet dissolved in methanol. Only colonies containing more than 50 cells were counted. The results were reported as the mean number of colonies observed in five randomly chosen microscope fields.

SA b-gal staining is widely used as a biomarker of cellular senescence in vivo and in vitro, with the positive green or blue-colored staining of b-galactosidase at pH 6.0 being remarkably increased in senescent cells. Senescent cells in frozen tissues or HCC cell lines were analyzed using a SA b-gal staining kit (Beyotime Inc., Nantong, China) according to the manufacturer’s instructions. The percentage of SA-b-gal positive cells was calculated by counting the cells in 5 random fields (at least 100 cells) using bright-field microscopy. The staining results for HCC samples were recorded as ‘‘positive’’ or ‘‘negative’’, according to the method reported by te Poele et al. [16].

Total RNA was isolated from cells using the RNAfast200 Kit (Fastagen Biotech, Shanghai, China). Reverse transcription was performed using the PrimeScriptÒ RT reagent Kit (TaKaRa Biotechnology, Dalian, China). The mRNA expression was assayed in triplicate and normalized to the b-actin mRNA expression. The relative levels were calculated using the Comparative-Ct Method (DDCt method). The following primers were used for qRT–PCR: FoxM1:50 -AACCGCTACTTGACATTGG30 (sense) and 50 -GCAGTGGCTTCATCTTCC-30 (antisense); b-actin: 50 -ATCGTGCG TGTGACATTAAGGAG-30 (sense) and 50 -AGGAAGGAAGGCTGGAAGAGTG-30 (antisense); p16INK4a: 50 -CATCGCGATGTCGCACGGTA-30 (sense) and 50 -TACGAAAGCGG GGTGGGTTGTG-30 (antisense); p53: 50 -ATGAGCCGCCTGAGGTTGG-30 (sense) and 50 CAGCCTGGGCATCCTTGAGT-30 (antisense); p21: 50 -TGGCACCTCACCTGCTCTG30 (sense) and 50 -GTTTGGAGTGGTAGAAATCTGTCAT-30 (antisense). All primer pairs were synthesized by TaKaRa.

2.8. Immunofluorescence and immunohistochemical assays To examine the expressions of FoxM1 and p53, the cells or tissues were firstly fixed and blocked, and then incubated with anti-FoxM1 and anti-p53 monoclonal antibodies (Santa Cruz Biotechnology, CA, USA). For immunofluorescence assay, samples were incubated with FITC (green)- or rhodamine (red)-conjugated secondary antibodies (1:1000; Pierce, Rockford, IL, USA), followed by observation with a laser scanning microscope (Olympus, Tokyo, Japan). Similarly for immunohistochemical assay, slides were incubated with biotinlabel goat anti-mouse or anti-rabbit IgG, followed by horseradish peroxidase (HRP) to label streptavidin. The intensity of immunohistochemical staining was scored as 0 (negative), 1 (weak), 2 (moderate strong) or 3 (strong). The extent of staining was assessed based on the percentage of positive tumor cells: 0 (negative), 1 (1–25%), 2 (26–50%), 3 (51–75%), and 4 (76–100%). The final staining score for each sample was the mean of the sum of the intensity and extent scores from three fields. The expression was considered as low if the final score was 1–5 and as high if the final score was 6–12.

2.9. TUNEL assay Apoptosis in HCC sections was measured and quantified with the use of the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay. Tumor sections were stained with TUNEL agent (Roche, Shanghai, China) following the manufacture’s instructions. Tissue sections included in the kit were stained and served as positive controls. The TUNEL-positive cells were counted under 5 randomly selected 400 microscopic fields. The apoptotic index was calculated as: the number of apoptotic cells/total number of nucleated cells  100%.

2.10. Western blot assay Cells were lysed in RIPA buffer (Beyotime Inc., NanTong, China). Protein concentration was determined with the Bradford reagent (Beyotime Inc.). Equal amounts of total proteins were separated and transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA). The membranes were subsequently immunoblotted with the appropriate primary antibody at 4 °C for 12 h, and then incubated with HRP conjugated anti-goat or anti-rabbit antibody (Santa Cruz). Signals were detected using the ECL Kit (Pierce, Rockford, IL).

2.11. Statistical analysis 2.5. Reactive oxygen species (ROS) assay DCFH-DA was cleaved intracellularly by nonspecific esterases and transformed to highly fluorescent 2,7-dichlorofluorescein (DCF) upon oxidation by ROS. After adding carboxy-DCFDA at a final concentration of 10 lM to the culture medium, the cells were incubated at 37 °C for an additional 30 min, harvested, washed with PBS, and immediately measured by FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA, USA).

All data were expressed as mean ± standard error of measurement (SEM) and analyzed by SPSS 11.0 software (SPSS Inc., Chicago, IL, USA). Categorical data were analyzed by Fisher’s exact test. The cumulative survival and recurrence rates were estimated by using the Kaplan–Meier method and the log-rank test. Comparisons of quantitative data between two groups were analyzed by Student’s t test, with P < 0.05 or P < 0.01 considered significant. Mean values of three independent experiments were presented for all samples.

107

K. Qu et al. / Cancer Letters 331 (2013) 105–114 Table 1 Correlation between FoxM1 expression and clinicopathological characteristics of HCC patients. Clinicopathological variables

Systemic L-OHP chemotherapy (Total N = 49) FoxM1 low (n = 18)

FoxM1 high (n = 31)

Gender Male Female

15 (83.3) 3 (16.7)

24 (77.4) 7 (22.6)

Age (yrs) <50 >50

4 (38.9) 14 (61.1)

11 (35.5) 20 (64.5)

TNM stage I II III IV

0 0 8 (44.4) 10 (55.6)

0 0 11 (35.5) 20 (64.5)

Differentiation I–II III–IV

8 (44.4) 10 (55.6)

5 (16.1) 26 (83.9)

HBsAg Positive Negative

15 (83.3) 3 (16.7)

24 (77.4) 7 (22.6)

Serum AFP <400 lg/L P400 lg/L

10 (55.6) 8 (44.4)

12 (38.7) 19 (61.3)

Maximal tumor size 65 cm >5 cm

4 (22.2) 14 (77.8)

16 (51.6) 15 (48.4)

Tumor number Single Multiple

15 (83.3) 3 (16.7)

14 (45.2) 17 (54.8)

Tumor encapsulation Absent Present

8 (44.4) 10 (55.6)

18 (58.1) 13 (41.9)

Recurrence Yes No

9 (50.0) 9 (50.0)

23 (74.2) 8 (25.8)

Hepatic arterial infusion with L-OHP (Total N = 42) P value

FoxM1 low (n = 27)

FoxM1 high (n = 15)

20 (74.1) 7 (25.9)

11 (73.3) 4 (26.7)

8 (29.6) 19 (70.4)

6 (40.0) 9 (60.0)

25 (92.6) 2 (7.4) 0 0

9 (60.0) 2 (13.3) 4 (26.7) 0

7 (25.9) 20 (74.1)

8 (53.3) 7 (46.7)

23 (85.2) 4 (14.8)

14 (93.3) 1 (6.7)

19 (70.4) 8 (29.6)

10 (66.7) 5 (33.3)

17 (63.0) 10 (37.0)

4 (26.7) 11 (73.3)

24 (88.9) 3 (11.1)

11 (73.3) 4 (26.7)

17 (63.0) 10 (37.0)

13 (86.7) 2 (13.3)

5 (18.5) 22 (81.5)

4 (26.7) 11 (73.3)

0.694

P value 0.958

0.873

0.495

0.535a

0.012b,*

0.030*

0.076

0.620

0.639

0.253

0.804

0.044*

0.024*

0.009**

0.195

0.357

0.158

0.086

0.698

a

Fisher’s exact test was performed between III stage and IV stage. Fisher’s exact test was performed between I, II stage and III stages. * P < 0.05. ** P < 0.01. b

3. Results 3.1. FoxM1 overexpression in hepatocellular carcinoma was associated with therapeutic response to oxaliplatin The western blotting of HCC samples showed that the FoxM1 overexpression were exclusive in HCCs (Fig. 1A). To further confirm these data, we evaluated the mRNA expression of FoxM1 by qRT– PCR in sixteen pairs of HCC and adjacent non-HCC liver samples. We found that the mRNA expression of FoxM1 in HCC samples were 2.5 fold higher than adjacent non-HCC liver samples (P < 0.01, Fig. 1B). This relationship was also experimentally verified in our HCC cell lines (HepG2, SMMC-7721, MHCC-97H, MHCC-97L and Hep3B), with FoxM1 up-regulated in four (HepG2, SMMC-7721, MHCC-97H and MHCC-97L) of the cell lines (Fig. 1C and D). Several studies have previously reported that FoxM1 overexpression is related to poor prognosis in HCC patients [17]. In our study, we also observed that the overexpression of FoxM1 impaired the response to oxaliplatin treatment in 49 advanced HCC patients. The follow-up analyses indicated that those patients carrying a high expression of FoxM1 in tumor had a shorter overall survival time and obvious earlier tumor recurrence than the patients with a low expression of FoxM1. The median overall survival and recurrence-free survival were 9.0 (confidence interval [95%CI] 6.1–11.9) and 5.5 months (95% CI 4.0–6.9) in the high FoxM1

expression group, while 14.2 (95% CI 8.0–20.5) and 12.5 (95% CI 3.6–21.4) in the low FoxM1 expression group (P < 0.05, Fig. 1E). 3.2. Cellular senescence induced by oxaliplatin was correlated with FoxM1 expression Oxaliplatin inhibited the growth of HepG2 and SMMC-7721 cells in a dose- and time-dependent manner, as demonstrated by MTT and colony formation assays (Fig. 2A and B). When exposed to oxaliplatin, HepG2 and SMMC-7721 cells were accumulated in the G1 phase, which is related to intracellular ROS generation, introducing oxaliplatin as a potential agent for induction of cellular senescence (Fig. 2C and D). To further explore the cellular senescence induced by oxaliplatin, we analyzed the phenotypes of HepG2 and SMMC-7721 cells according to the SA b-gal staining, which is the most widely used biomarker for senescent and aging cells. We found that the positive rate of the SA b-gal staining varied with the oxaliplatin concentration (1.25, 2.5 and 5.0 lg/mL), with the 2.5 lg/mL oxaliplatin group exhibiting the highest senescence percent in a time-dependent manner (P < 0.05, Fig. 3A and B). We further performed the SA b-gal staining on HCC samples resected from the 42 patients who received HAI with oxaliplatin and on the 25 controls. Interestingly, we found that the positive SA b-gal staining rate was significantly higher in the patients treated with oxaliplatin (P < 0.05, Fig. 3E), and meanwhile, the overexpression of FoxM1 was nega-

108

K. Qu et al. / Cancer Letters 331 (2013) 105–114

Fig. 1. FoxM1 overexpression in hepatocellular carcinoma correlated with lack of response to oxaliplatin. (A) and (B) paired HCC tissues were collected to evaluate FoxM1 by western blot and RT–PCR, which was quantified using b-actin as normalization control. (C) and (D) FoxM1 expressions in immortal hepatocytes (L02) and five HCC cell lines (HepG2, SMMC-7721, MHCC-97H, MHCC-97L and Hep3B) were evaluated at protein and mRNA levels. (E) and (F) Kaplan–Meier analysis of the overall survival and recurrence probability of indicated HCC patients treated with oxaliplatin. Data are represented as mean ± SEM from three independent experiments. P < 0.01.

tively related to SA b-gal staining (P < 0.05, Fig. 3D–F). However, there was no significant relationship between the apoptosis index of oxaliplatin treated patients and the FoxM1 expression, nor between the apoptosis index and the SA b-gal staining rate (P < 0.05, Fig. 3C and G). These data indicated that reduced expression of FoxM1 was involved in the cellular senescence induced by oxaliplatin.

3.3. Oxaliplatin-induced cellular senescence required activation of p53 and p21, and repression of FoxM1 To elucidate the mechanism of oxaliplatin-induced senescence in HCC cells, we treated HepG2 and SMMC-7721 cells with 2.5 lg/mL oxaliplatin for 120 h. Cells were collected every 24 h for evaluation by qRT–PCR. As has been reported that altered expressions of p53, p21 and p16INK4a would contribute to senescence pathway, we also found that oxaliplatin treatment upregulated p53 and p21 mRNA levels, whereas the p16INK4a level remained constant [7,18] (Fig. 4A). Silencing of p53 significantly decreased oxaliplatin induced cellular senescence in both HepG2 and SMMC-7721 cells, suggesting that p53 played a critical role in oxaliplatin-induced cellular senescence (Fig. 4B).

In addition to increasing cellular senescence, oxaliplatin also resulted in the decrease in the mRNA level of FoxM1 (Fig. 4C). Similarly, results of western blotting also confirmed the decreased protein level of FoxM1 after treatment with 2.5 lg/mL oxaliplatin for 120 h. On the contrary, during oxaliplatin exposure, the protein levels of p53 and p21 increased uniformly and reached a maximum at 120 h (Fig. 4D). Immunofluorescence analysis also indicated that the FoxM1 protein expression was significantly downregulated in the nuclei of HCC cells after oxaliplatin treatment, whereas the wild-type p53 expression was upregulated (Fig. 4E).

3.4. FoxM1 was down-regulated by p53 but not by p21 in oxaliplatin treatment P53 represses the transcription of a substantial number of genes, among which, FoxM1 was identified by three independent studies as significantly downregulated upon p53 activation in breast cancer [19–21]. However, it is still unknown whether FoxM1 is affected by p53 when it is involved in the senescence of HCC cells. In the present study, when oxaliplatin-treated HepG2 and SMMC-7721 cells were transfected with siRNA-p53, the mRNA and protein levels of FoxM1 were both up-regulated, while those of

K. Qu et al. / Cancer Letters 331 (2013) 105–114

109

Fig. 2. Effects of oxaliplatin on the growth, cell cycle and ROS level in HepG2 and SMMC-7721 cells. (A) HCC cells treated with different doses of oxaliplatin for 48 h were recultured in drug-free medium for 72 h and examined by MTT assay. (B) Colony formation analysis indicated that HCC cells were significantly inhibited by oxaliplatin. (C) Flow cytometry showed that cells were accumulated in the G1 phase after 48 h oxaliplatin treatment. (D) ROS level of HCC cells after 48 h oxaliplatin treatment. Data are represented as mean ± SEM from three independent experiments. P < 0.05, P < 0.01; L-OHP: oxaliplatin.

p21 were down-regulated (Fig. 5A and B). Surprisingly, although silencing of the p53 expression abrogated the induction of p21 and up-regulated FoxM1, silencing of p21 had little effects on the upregulation of FoxM1, suggesting that oxaliplatin repressed the FoxM1 expression via p53-dependent and p21-independent pathways (Fig. 5C and D). 3.5. FoxM1 countered cellular senescence via regulating the expressions of cycle related proteins p21 and p27, and cyclins D1 and B1 The highest level of senescence with obvious G1 phase arrest was observed in the siRNA-FoxM1 group when the FoxM1 expression was mediated by transfecting HepG2 and SMMC-7721 cells with siRNA-FoxM1 and pcDNA 3,1-FoxM1 (Fig. 6A and B). Since FoxM1 is able to transcriptionally upregulate the expression of

ubiquitin ligase Skp2 and Cks1, which are specific targets of these cyclin-dependent kinase inhibitors (CDKIs) for degradation [22], we verified the relationship between the expressions of FoxM1 and the CDKI protein p21 and p27 (Fig. 6C). We found that p21 and p27 were downregulated due to FoxM1 overexpression. On the other hand, cyclins D1 and B1, as cycle positive regulators which could be inhibited by CDKIs, were positively regulated by FoxM1. However, the modulation of FoxM1 had little effect on the mRNA and protein expression of p53 and p19ARF which negatively regulates p53 (Fig. 6C and D). 3.6. FoxM1 overexpression impaired the oxaliplatin-induced senescence of HCC cells Since FoxM1 played a critical role in oxaliplatin-induced senescence, we further investigated whether silencing of FoxM1

110

K. Qu et al. / Cancer Letters 331 (2013) 105–114

Fig. 3. Senescence was induced by oxaliplatin in HepG2 and SMMC-7721 cells. (A) Senescence analysis of HCC cells treated with different doses of oxaliplatin, with the highest positive SA b-gal staining in 2.5 lg/mL treatment for 72 h, scale bar, 100 lm. (B) Positive SA b-gal staining of the HCC cells treated with 2.5 lg/mL oxaliplatin from 24 h to 120 h. (C) FoxM1, SA b-gal and TUNEL staining of the HCC tissues from patients treated by HAI with oxaliplatin, scale bar, 150 lm. (D) Correlation analysis of FoxM1 expression and oxliplatin treatment in the HCC, with oxaliplatin treatment in 42 cases and no oxaliplatin treatment in 25 cases. (E) Correlation analysis of cellular senescence and oxliplatin treatment, represented by SA b-gal staining. (F) Correlation analysis of cellular senescence and FoxM1 expression in the 42 HCC tissues treated with oxaliplatin. (G) Apoptosis index analysis in the 42 HCC patients grouped by FoxM1 expression and senescence, represented by TUNEL assay. Data are represented as mean ± SEM from three independent experiments. P < 0.05 and P < 0.01 vs. control in HepG2 cells; #P < 0.05 and ##P < 0.01 vs. control in SMMC-7721 cells.

combined with oxaliplatin treatment could accelerate cellular senescence. Our results revealed that silencing of FoxM1 combined with oxaliolatin treatment increased the G1 phase arrest and promoted senescence in HepG2 and SMMC-7721 cells. On the other hand, it showed a prominent anti-senescence effect, as demonstrated by the FoxM1 up-regulation (Fig. 7A and B). In summary, these results indicated that FoxM1 was involved in the mediation of senescence, which determines the response to oxaliplatin treatment. 4. Discussion Since its initial discovery by Hayflick et al., cellular senescence, which could induce irreversible arrest of cell division, has been shown to exist in various mammalian tissues [23]. Recent studies have demonstrated that chemotherapy-induced senescence is one of the key determinants of tumor response to therapy, since chemotherapy drugs, especially those affecting cell cycle in mitosis and DNA replication, can induce senescence-like morphological changes in tumor cells [16,24]. DNA damage induced by platinum compounds is commonly thought to be mediated by various cytotoxic effects, including apoptosis, autophagy and senescence.

Specifically, cisplatin, the first platinum drug, has been demonstrated to induce cell senescence in various tumor cell lines [25,26]. Oxaliplatin, with pharmacological effects similar to cisplatin, has also been shown in preclinical studies to exert cytotoxic effects in tumor cells. However, up to date, no studies have been done to investigate the cellular senescence induced by oxaliplatin. In this study, we demonstrated the significant growth inhibition and cell cycle arrest after oxaliplatin treatment. In addition, our results also verified the increased intracellular ROS induced by oxaliplatin treatment in HCC [27]. Considering oxidative stress as an important factor inducing senescence, we successfully revealed by using SA b-gal staining that senescence was induced in oxaliplatin-treated cells in a dose- and time-dependent manner. Furthermore, similar results were found in HCC tissues. We found that senescence was relatively more prominent at less cytotoxic drug doses, suggesting the possibility of continuous infusion of oxaliplatin in patients’ plasma [28]. Therefore, cellular senescence may be a significant determinant of tumor response to oxaliplatin in HCC patients [7]. Growth-regulatory genes such as ATM, Bmi-1 and PTEN have been reported to be involved in chemo-induced senescence

K. Qu et al. / Cancer Letters 331 (2013) 105–114

111

Fig. 4. Oxaliplatin-induced senescence was triggered by activation of p53 and repression of FoxM1. (A) HCC cells were treated with 2.5 lg/mL oxaliplatin from 24 h to 120 h. The expression of p53, p21 and p16INK4a were evaluated by real-time PCR. p53 and p21 were increased while p16INK4a remained constant. (B) Senescence analysis of HCC cells transfected with siRNA-p53. Silencing of p53 significantly inhibited senescence induced by oxaliplatin. (C) The decreasing expression of FoxM1 during the 120 h exposure to 2.5 lg/mL oxaliplatin was detected by real-time PCR. (D) Western blot analysis of FoxM1, wild-type p53, p21 and p16INK4a in HCC cells treated by oxaliplatin. FoxM1 was decreased during oxaliplatin exposure, which was inversely correlated with p53 and p21 alterations. (E) HCC cells treated with 2.5 lg/mL oxaliplatin for 72 h were stained by FoxM1 (red) and wild-type p53 (green) antibodies. Images were visualized by confocal microscopy. Data are represented as mean ± SEM from three independent experiments. P < 0.05 and P < 0.01 vs. control in HepG2 cells; #P < 0.05 and ##P < 0.01 vs. control in SMMC-7721 cells. Scale bars, 100 lm.

[29–31]. Research on senescence-related genes may, therefore, help to elucidate the mechanism of chemo-related senescence. FoxM1, a critical regulator of cell cycle progression, has been found

to participate in cell proliferation, onconogenesis, aging and tumorigenesis [9,32–36]. Thus, FoxM1-depleted cells may have difficulty executing mitosis and may exhibit chromosomal

112

K. Qu et al. / Cancer Letters 331 (2013) 105–114

Fig. 5. FoxM1 expression was up-regulated by silencing of wild-type p53, but not by silencing of p21 in oxaliplatin treated HCC cells. (A) and (B) HCC cells transfected with siRNA-p53 were treated with 2.5 lg/mL oxaliplatin for 96 h. qRT–PCR and western blot analysis of FoxM1 expression were performed in the treated HCC cells. Silencing of p53 expression significantly down-regulated p21 and up-regulated FoxM1. (C) and (D) qRT–PCR and western blot analyses of FoxM1 expression in the HCC cells transfected with siRNA-p21. Data are represented as mean ± SEM from three independent experiments. P < 0.05, P < 0.01. Scale bars, 150 lm.

Fig. 6. Down-regulation of FoxM1 could trigger senescence via counteracting inhibition of p21 and p27 and down-regulating cyclins D1 and B1, but it had little influence on p53 and p19ARF expression. (A) Senescence analysis was performed in the control, mock, siFoxM1, pcDNA 3.1 and pcDNA 3.1-FoxM1 groups by SA b-Gal staining. (B) G1 phase arrest was detected in the siRNA-FoxM1 transfected cells by flow cytometry. (C) Western blot analysis of p53, p19ARF, p21, p27, cyclins D1 and B1 in HCC cells transfected with siRNA or pcDNA 3.1-FoxM1 for 24 h. (D) qRT–PCR analysis showed that silencing of FoxM1 did not significantly down-regulate p53 and p19ARF mRNA expression. Data are represented as mean ± SEM from three independent experiments. P < 0.05, P < 0.01.

instability and polyploidy [22,37]. In our study, the expression of FoxM1 was significantly correlated with G1 arrest and cellular senescence in HepG2 and SMMC-7721 cells, indicating that the increased FoxM1 expression in HCC cells would exert antagonistic effect against oxaliplatin-induced senescence. In addition, although FoxM1 expression was significantly correlated with SA b-gal staining, but it had no obvious relationship with the apoptosis index of the HCC tissues receiving oxaliplatin treatment. P53 has been widely recognized as a regulator of senescence in various tumors and it may promote tumor inhibition by inducing

senescence [31,38]. In the present study, we confirmed and extended these findings in HCC cell lines and demonstrated that FoxM1 was inversely regulated by p53 in oxaliplatin-induced senescence. In addition, p21, as a p53-regulated CDKI protein, was also involved in cellular damage caused by drug-induced growth arrest. On the basis of our current findings that DNA damage induced by cisplatin might induce irreversible growth arrest involving p53 and p21, our results also revealed that up- and down-regulation of FoxM1 caused a remarkable post-translational change in the expression of p21, whereas the expression of p53

K. Qu et al. / Cancer Letters 331 (2013) 105–114

113

Fig. 7. Effects of oxaliplatin (2.5 lg/mL for 72 h) plus FoxM1 down-regulation on cellular senescence. (A) Cell cycle analysis of HCC cells undergoing the indicated treatment by flow cytometry. Cells were increased in G1 phase. (B) Senescence analysis of HCC cells undergoing the indicated treatment. The combination treatment showed the best effect. Data are represented as mean ± SEM from three independent experiments. P < 0.05, P < 0.01.

molecules FoxM1 and p21, while alteration of p21 and FoxM1 expression has little effect on p53, suggesting the existence of specific molecular pathway ‘‘p53-FoxM1-p21’’ and that oxaliplatin represses FoxM1 expression via p53-dependent pathways [39]. We also found that p27, cyclinB1 and cyclinD1, which are cycle related proteins involved in cellular damage, were affected by the regulation of FoxM1, further confirming the involvement of FoxM1 in the cellular senescence in HCC [40–42] (Fig. 8). As FoxM1 appears to be associated with poor prognosis of HCC patients receiving oxaliplatin therapy because of its antisenescence activity, we call for the development and optimization of FoxM1 inhibitors. Silencing of FoxM1 by small RNA interference significantly promoted the senescence induced by oxaliplatin, while overexpressed FoxM1 resulted in inhibited senescence, which partially elucidated the FoxM1 mediation of the therapeutic response to oxaliplatin in HCC patients. However, further studies are necessary to evaluate the full potential of FoxM1 inhibition in oxaliplatin therapy and to determine its safe application in the promotion of cellular senescence and chemotherapeutic sensitivity. Acknowledgments This work was financially supported by Grants from the National Natural Science Foundation of China (Nos. 30872482, 81072051, 81201549 and 81272644). References Fig. 8. Overview of FoxM1 related pathways for oxaliplatin-induced senescence in HCC cells. Free from drug treatment, proliferation of HCC cells depended on the activation of FoxM1. When exposed to oxaliplatin, the p53 activated by ROS would inhibit FoxM1 expression, resulting in the up-regulation of CDK inhibitors (p27 and p21) and the downregulation of cycle positive regulators (cyclins D1 and B1). The disturbed balance of cycle-related proteins would cause cell cycle arrest and eventually induce senescence.

was hardly changed. On the other hand, according to our previous results, modulation of p53 can significantly affect its downstream

[1] D.M. Parkin, F. Bray, J. Ferlay, P. Pisani, Global cancer statistics, CA Cancer J. Clin. 55 (2005) (2002) 74–108. [2] C.B. O’Suilleabhain, R.T. Poon, J.L. Yong, G.C. Ooi, W.K. Tso, S.T. Fan, Factors predictive of 5-year survival after transarterial chemoembolization for inoperable hepatocellular carcinoma, Brit. J. Surg. 90 (2003) 325–331. [3] A. Burroughs, D. Hochhauser, T. Meyer, Systemic treatment and liver transplantation for hepatocellular carcinoma: two ends of the therapeutic spectrum, Lancet Oncol. 5 (2004) 409–418. [4] Y. Yen, D.W. Lim, V. Chung, R.J. Morgan, L.A. Leong, S.I. Shibata, L.D. Wagman, H. Marx, P.G. Chu, J.A. Longmate, H.J. Lenz, R.K. Ramanathan, C.P. Belani, D.R.

114

[5] [6]

[7] [8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18] [19] [20]

[21] [22]

[23]

[24]

K. Qu et al. / Cancer Letters 331 (2013) 105–114 Gandara, Phase II study of oxaliplatin in patients with unresectable, metastatic, or recurrent hepatocellular cancer: a California Cancer Consortium Trial, Am. J. Clin. Oncol. 31 (2008) 317–322. L. Kelland, The resurgence of platinum-based cancer chemotherapy, Nat. Rev. Cancer 7 (2007) 573–584. D.A. Gewirtz, S.E. Holt, L.W. Elmore, Accelerated senescence: an emerging role in tumor cell response to chemotherapy and radiation, Biochem. Pharmacol. 76 (2008) 947–957. M. Ozturk, A. Arslan-Ergul, S. Bagislar, S. Senturk, H. Yuzugullu, Senescence and immortality in hepatocellular carcinoma, Cancer Lett. 286 (2009) 103–113. C.C. Deocaris, N. Widodo, B.G. Shrestha, K. Kaur, M. Ohtaka, K. Yamasaki, S.C. Kaul, R. Wadhwa, Mortalin sensitizes human cancer cells to MKT-077-induced senescence, Cancer Lett. 252 (2007) 259–269. I.M. Kim, T. Ackerson, S. Ramakrishna, M. Tretiakova, I.C. Wang, T.V. Kalin, M.L. Major, G.A. Gusarova, H.M. Yoder, R.H. Costa, V.V. Kalinichenko, The Forkhead Box m1 transcription factor stimulates the proliferation of tumor cells during development of lung cancer, Cancer Res. 66 (2006) 2153–2161. T.V. Kalin, I.C. Wang, T.J. Ackerson, M.L. Major, C.J. Detrisac, V.V. Kalinichenko, A. Lyubimov, R.H. Costa, Increased levels of the FoxM1 transcription factor accelerate development and progression of prostate carcinomas in both TRAMP and LADY transgenic mice, Cancer Res. 66 (2006) 1712–1720. Z. Wang, S. Banerjee, D. Kong, Y. Li, F.H. Sarkar, Down-regulation of Forkhead Box M1 transcription factor leads to the inhibition of invasion and angiogenesis of pancreatic cancer cells, Cancer Res. 67 (2007) 8293–8300. M. Liu, B. Dai, S.H. Kang, K. Ban, F.J. Huang, F.F. Lang, K.D. Aldape, T.X. Xie, C.E. Pelloski, K. Xie, R. Sawaya, S. Huang, FoxM1B is overexpressed in human glioblastomas and critically regulates the tumorigenicity of glioma cells, Cancer Res. 66 (2006) 3593–3602. S.K. Li, D.K. Smith, W.Y. Leung, A.M. Cheung, E.W. Lam, G.P. Dimri, K.M. Yao, FoxM1c counteracts oxidative stress-induced senescence and stimulates Bmi1 expression, J. Biol. Chem. 283 (2008) 16545–16553. H.J. Park, J.R. Carr, Z. Wang, V. Nogueira, N. Hay, A.L. Tyner, L.F. Lau, R.H. Costa, P. Raychaudhuri, FoxM1, a critical regulator of oxidative stress during oncogenesis, EMBO J. 28 (2009) 2908–2918. J. Zeng, L. Wang, Q. Li, W. Li, M. Bjorkholm, J. Jia, D. Xu, FoxM1 is up-regulated in gastric cancer and its inhibition leads to cellular senescence, partially dependent on p27 kip1, J. Pathol. 218 (2009) 419–427. R.H. te Poele, A.L. Okorokov, L. Jardine, J. Cummings, S.P. Joel, DNA damage is able to induce senescence in tumor cells in vitro and in vivo, Cancer Res. 62 (2002) 1876–1883. L. Xia, W. Huang, D. Tian, H. Zhu, Y. Zhang, H. Hu, D. Fan, Y. Nie, K. Wu, Upregulated FoxM1 expression induced by hepatitis B virus X protein promotes tumor metastasis and indicates poor prognosis in hepatitis B virus-related hepatocellular carcinoma, J. Hepatol. 57 (2012) 600–612. S.H. Park, J.S. Lim, K.L. Jang, All-trans retinoic acid induces cellular senescence via upregulation of p16, p21, and p27, Cancer Lett. 310 (2011) 232–239. B. Pandit, M. Halasi, A.L. Gartel, P53 negatively regulates expression of FoxM1, Cell Cycle 8 (2009) 3425–3427. J. Millour, N. de Olano, Y. Horimoto, L.J. Monteiro, J.K. Langer, R. Aligue, N. Hajji, E.W. Lam, ATM and p53 regulate FOXM1 expression via E2F in breast cancer epirubicin treatment and resistance, Mol. Cancer Ther. 10 (2011) 1046–1058. A.M. Barsotti, C. Prives, Pro-proliferative FoxM1 is a target of p53-mediated repression, Oncogene 28 (2009) 4295–4305. I.C. Wang, Y.J. Chen, D. Hughes, V. Petrovic, M.L. Major, H.J. Park, Y. Tan, T. Ackerson, R.H. Costa, Forkhead box M1 regulates the transcriptional network of genes essential for mitotic progression and genes encoding the SCF (Skp2Cks1) ubiquitin ligase, Mol. Cell Biol. 25 (2005) 10875–10894. G.P. Dimri, X. Lee, G. Basile, M. Acosta, G. Scott, C. Roskelley, E.E. Medrano, M. Linskens, I. Rubelj, O. Pereira-Smith, et al., A biomarker that identifies senescent human cells in culture and in aging skin in vivo, Proc. Natl. Acad. Sci. USA 92 (1995) 9363–9367. B.D. Chang, E.V. Broude, M. Dokmanovic, H. Zhu, A. Ruth, Y. Xuan, E.S. Kandel, E. Lausch, K. Christov, I.B. Roninson, A senescence-like phenotype distinguishes tumor cells that undergo terminal proliferation arrest after exposure to anticancer agents, Cancer Res. 59 (1999) 3761–3767.

[25] K. Fang, C.C. Chiu, C.H. Li, Y.T. Chang, H.T. Hwang, Cisplatin-induced senescence and growth inhibition in human non-small cell lung cancer cells with ectopic transfer of p16INK4a, Oncol. Res. 16 (2007) 479–488. [26] X. Wang, S.C. Wong, J. Pan, S.W. Tsao, K.H. Fung, D.L. Kwong, J.S. Sham, J.M. Nicholls, Evidence of cisplatin-induced senescent-like growth arrest in nasopharyngeal carcinoma cells, Cancer Res. 58 (1998) 5019–5022. [27] S.C. Lim, J.E. Choi, H.S. Kang, S.I. Han, Ursodeoxycholic acid switches oxaliplatin-induced necrosis to apoptosis by inhibiting reactive oxygen species production and activating p53-caspase 8 pathway in HepG2 hepatocellular carcinoma, Int. J. Cancer 126 (2010) 1582–1595. [28] W.P. Petros, G. Broadwater, D. Berry, R.B. Jones, J.J. Vredenburgh, C.J. Gilbert, J.P. Gibbs, O.M. Colvin, W.P. Peters, Association of high-dose cyclophosphamide, cisplatin, and carmustine pharmacokinetics with survival, toxicity, and dosing weight in patients with primary breast cancer, Clin. Cancer Res. 8 (2002) 698–705. [29] J.J. Jacobs, K. Kieboom, S. Marino, R.A. DePinho, M. van Lohuizen, The oncogene and Polycomb-group gene bmi-1 regulates cell proliferation and senescence through the ink4a locus, Nature 397 (1999) 164–168. [30] U. Herbig, W.A. Jobling, B.P. Chen, D.J. Chen, J.M. Sedivy, Telomere shortening triggers senescence of human cells through a pathway involving ATM, p53, and p21(CIP1), but not p16(INK4a), Mol. Cell 14 (2004) 501–513. [31] Z. Chen, L.C. Trotman, D. Shaffer, H.K. Lin, Z.A. Dotan, M. Niki, J.A. Koutcher, H.I. Scher, T. Ludwig, W. Gerald, C. Cordon-Cardo, P.P. Pandolfi, Crucial role of p53dependent cellular senescence in suppression of Pten-deficient tumorigenesis, Nature 436 (2005) 725–730. [32] H.C. Sun, M. Li, J.L. Lu, D.W. Yan, C.Z. Zhou, J.W. Fan, X.B. Qin, H.M. Tang, Z.H. Peng, Overexpression of Forkhead box M1 protein associates with aggressive tumor features and poor prognosis of hepatocellular carcinoma, Oncol. Rep. 25 (2011) 1533–1539. [33] P. Raychaudhuri, H.J. Park, FoxM1: a master regulator of tumor metastasis, Cancer Res. 71 (2011) 4329–4333. [34] H.J. Park, G. Gusarova, Z. Wang, J.R. Carr, J. Li, K.H. Kim, J. Qiu, Y.D. Park, P.R. Williamson, N. Hay, A.L. Tyner, L.F. Lau, R.H. Costa, P. Raychaudhuri, Deregulation of FoxM1b leads to tumour metastasis, EMBO Mol. Med. 3 (2011) 21–34. [35] M. Frau, F. Biasi, F. Feo, R.M. Pascale, Prognostic markers and putative therapeutic targets for hepatocellular carcinoma, Mol. Aspects Med. 31 (2010) 179–193. [36] J. Laoukili, M.R. Kooistra, A. Bras, J. Kauw, R.M. Kerkhoven, A. Morrison, H. Clevers, R.H. Medema, FoxM1 is required for execution of the mitotic programme and chromosome stability, Nat. Cell Biol. 7 (2005) 126–136. [37] Q.F. Wu, C. Liu, M.H. Tai, D. Liu, L. Lei, R.T. Wang, M. Tian, Y. Lu, Knockdown of FoxM1 by siRNA interference decreases cell proliferation, induces cell cycle arrest and inhibits cell invasion in MHCC-97H cells in vitro, Acta Pharmacol. Sin. 31 (2010) 361–366. [38] C.A. Schmitt, J.S. Fridman, M. Yang, S. Lee, E. Baranov, R.M. Hoffman, S.W. Lowe, A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy, Cell 109 (2002) 335–346. [39] B.D. Chang, Y. Xuan, E.V. Broude, H. Zhu, B. Schott, J. Fang, I.B. Roninson, Role of p53 and p21waf1/cip1 in senescence-like terminal proliferation arrest induced in human tumor cells by chemotherapeutic drugs, Oncogene 18 (1999) 4808– 4818. [40] I. Kikuchi, Y. Nakayama, T. Morinaga, Y. Fukumoto, N. Yamaguchi, A decrease in cyclin B1 levels leads to polyploidization in DNA damage-induced senescence, Cell Biol. Int. 34 (2010) 645–653. [41] P.K. Majumder, C. Grisanzio, F. O’Connell, M. Barry, J.M. Brito, Q. Xu, I. Guney, R. Berger, P. Herman, R. Bikoff, G. Fedele, W.K. Baek, S. Wang, K. Ellwood-Yen, H. Wu, C.L. Sawyers, S. Signoretti, W.C. Hahn, M. Loda, W.R. Sellers, A prostatic intraepithelial neoplasia-dependent p27 Kip1 checkpoint induces senescence and inhibits cell proliferation and cancer progression, Cancer Cell 14 (2008) 146–155. [42] W.H. Lien, C.K. Chen, L.Y. Lai, Y.H. Chen, M.P. Wu, L.W. Wu, Participation of cyclin D1 deregulation in TNP-470-mediated cytostatic effect: involvement of senescence, Biochem. Pharmacol. 68 (2004) 729–738.