Background
Ovarian cancer is the sixth most common cancer among women and it leads to the highest mortality per year than any other cancers of the female reproductive system [
1,
2]. Cisplatin derivatives are first-line chemotherapeutic agents used for treatment of ovarian cancer. However, chemoresistance remains a major hurdle to successful therapy and results in low five-year survival rates [
2,
3]. Previous studies have suggested that resistance may be due to reduced drug accumulation [
4], increased levels of glutathione and metallothionein [
5] and enhanced DNA repair [
6]. At present, it is widely accepted that the apoptotic response of cancer cells to chemotherapeutic drugs is the determining force for sensitivity to chemotherapy [
7,
8]. Many molecules, including XIAP [
3], MKP3 [
9], PI3K, Akt2 [
10,
11], PTEN [
12], P-glycoprotein [
13] and MDR [
14], have been reported to be involved in the regulation of apoptosis and in the complicated signaling network that determines the fate of cancer cells, i.e., either "death" or "survival." Though much progress has been made, our recent studies have made efforts to predict the response of cancer cells to chemotherapeutic agents before treatment and to identify the possible alterations that mediate resistance.
P21 was the first identified inhibitor of cyclin/cyclin-dependent kinase (CDK) complexes [
15]. Previous studies demonstrated that p21 could act as a "tumor suppressor" by binding to cellular CDK and proliferating cell nuclear antigen (PCNA), thereby inhibiting their function and leading to cell cycle arrest, leading to blockade of DNA synthesis and inhibition of cell proliferation [
15,
16]. Numerous studies that analyzed the expression of p21 in different types of human cancers have revealed that loss of p21 correlates with carcinogenesis and a poor prognosis in small-cell lung, colorectal, cervical and head and neck cancers [
17‐
20]. In contrast, other findings have found that increased p21 expression was associated with tumor progression in ovarian, cervical, breast and esophageal squamous cell carcinomas [
21‐
26]. This discrepancy could be due to the status of p21 itself and/or to differences in the histological types of cancers that have been analyzed. Besson et al. [
27] suggested that control of the subcellular localization of p21 could represent an important regulatory switch from a nuclear tumor suppressor to a cytoplasmic oncogene. Once phosphorylated by Akt, p21 is induced to emigrate from the nucleus to the cytoplasm, protecting cells from apoptosis [
28‐
31]. Additionally, clinical immunohistochemical analysis have proven that cytoplasmic p21 is a novel predictor of poor prognosis in breast cancer [
32,
33]. Although the role of p21 in the development of various types of human cancers has garnered much attention, little is known about its involvement in drug resistance.
Given that chemoresistance is a biological trait of tumor malignancy and has a direct influence on a patient's prognosis, this study was designed to explore whether cytoplasmic p21 correlated with cisplatin resistance in ovarian tumors. In addition, we sought to determine whether interfering with cytoplasmic p21 could enhance the susceptibility of cancer cells to cisplatin. Here, we report that in the cisplatin-resistant cell line C13*, p21 is predominantly localized to the cytoplasm, while in the cisplatin-sensitive cell line OV2008, p21 is mainly restricted to the nucleus. Additionally, the exposure to low-dose cisplatin in OV2008 induced translocation of most p21 protein from nuclear to cytoplasm. Knockdown of cytoplasmic p21 by transfection of p21 siRNA into C13* cells notably enhanced their sensitivity to cisplatin. Inhibition of p21 translocation into the cytoplasm by Akt2 shRNA transfection in C13* cells significantly enhanced their sensitivity as well. Conversely, the accumulation of p21 in the cytoplasm by transfection of active Akt2 in OV2008 conferred resistance to cisplatin in OV2008 cells. Further analysis of clinical ovarian tumor samples by immunohistochemistry revealed that p21 was predominantly localized in the nucleus in the drug sensitive group. In contrast, p21 is mainly localized in the cytoplasm in the drug resistant group.
Methods
Cell lines and cell culture
The cisplatin-resistant ovarian cancer cell line C13* and its parental variant OV2008 were provided by Doctor Benjamin K. Tsang (Department of Obstetrics, Gynecology and Cellular and Molecular Medicine, University of Ottawa, Canada). Cells were maintained in RPMI-1640 supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin and 10% fetal bovine serum (FBS) at 37°C in a humidified atmosphere containing 5% CO2.
Clinical samples
From 2004 through 2007, 40 patients who were diagnosed with ovarian carcinoma and received standard cisplatin-based intravenous chemotherapy after surgery at Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology were included in the study after obtainment of oral and written informed consents. Paraffin-embedded tumor tissue sections of each patient were prepared by the Department of Clinical Pathology. Information on histopathologic diagnosis was extracted from medical records and reviewed by a specialist in gynecologic pathology. The patients were selected and divided into treatment response and treatment non-response groups according to the CA125 criteria proposed by the Gynecological Cancer Intergroup (GCIC) [
34]. Briefly, treatment response patients were defined as having at least a 50% reduction in CA125 levels, when compared to pretreatment samples, and this reduction must have been maintained for at least 28 days; the intervening value of CA125 must have been less than or equal to the previous value. Patients demonstrated any clinical evidence of progression, such as increased lump size and ascites, were excluded from the treatment response group. The remaining patients were classified as treatment non-response group. This study was reviewed and approved by the ethics committee of the medical faculty at the Tongji Hospital, Tongji Medical College, Huazhong University of Science & Technology.
Antibodies and reagents
Antibodies against Akt, Ser473-phosphorylated Akt and p21 were purchased from Cell Signaling Technology, Inc. (Beverly, MA, USA). β-actin antibody and secondary goat anti-rabbit and goat anti-mouse alkaline phosphatase antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Anti-caspase 3 antibody was purchased from Biolegend Co. (San Diego, CA, USA). RPMI-1640, FBS, Lipofectamine 2000 and Trizol reagent were purchased from Invitrogen Co. (Carlsbad, CA, USA). Cisplatin, MTT, DMSO, G418 and DAPI dye were obtained from Sigma Chemical Co. (St. Louis, MO, USA). The NE-PER cytoplasmic protein extraction kit was purchased from Pierce Co. (Rockford, IL, USA). The annexin-V/PI apoptosis detection kit was purchased from Promoter Biotech Ltd. (Wuhan, Hubei, China).
Construction of plasmids
A constitutively active Akt expression vector (AAkt2) and short hairpin RNA targeting Akt2 (Akt2Sh) were described previously [
11,
35]. The sequences of the p21 RNA interfering fragment (p21si) were as follows: sense, 5'- CUU CGA CUU UGU CAC CGA G -3'; anti-sense, 5'- C UCG GUG ACA AAG UCG AAG -3' [
36]. The sequences of the mismatched fragment (p21sm) were: sense, 5'- CUC GAC UUC GUA CCC GAG -3': anti-sense, 5'- CUC GGG UAC GAA GUC GAG -3'.
Transient transfection for RNAi targeting
For RNAi targeting, C13* cells cultured in 6-well plates were transfected with indicated plasmids using Lipofectamine 2000. After 6 hours of incubation, the transfection solution was removed, and was replaced with fresh complete growth medium. 48 hours post-transfection the cells were assayed for the expression of p21 and treated with cisplatin for further experiment.
Establishment of stable-expression cell lines of AAkt2 in OV2008 cells and Akt2sh in C13* cells
OV2008 cancer cells were stably transfected with AAkt2 vector, C13* cancer cells were stably transfected with Akt2sh, using Lipofectamine 2000. Their corresponding empty vectors, i.e., pcDNA3.1 and pEGFPC1, were transfected as negative control. The cells were selected with G418. The concentration of G418 for selection and maintenance was 600 μg/μl. After three weeks the G418-resistant cell pools were established and seeded into 100 mm dishes for further propagation.
Real-time PCR
Total RNA was isolated from each group of cells using Trizol Regent, according to the manufacturer's instruction. Real-time PCR amplifications were carried out using DNase I (Promega)-treated total RNA. Reactions were performed in a Stratagene MX3000P system using the Real-time PCR Master Mix (TOYOBO, Japan). P21 primer sequences were as follows: sense, 5'-CCT CTT CGG CCC GGT GGA C-3': anti-sense, 5'-CCG TTT TCG ACC CTG AGA G-3'. GAPDH primer sequences were as follows: sense, 5'-ACG GAT TTG GTC GTA TTG GG-3'; anti-sense, 5'-TGA TTT TGG AGG GAT CTC GC-3'.
Western blot analysis
Total proteins were extracted by lysing cells in buffer containing 50 mM Tris pH 7.4, 150 mM NaCl, 0.5% NP-40, 50 mM NaF, 1 mM Na3VO4, 1 mM phenylmethylsulfonyl fluoride, 25 mg/ml leupeptin and 25 mg/ml aprotinin. The lysates were cleared by centrifugation, and the supernatants were collected. Cytoplasmic proteins were extracted using the N-PER cytoplasmic protein extraction kit according to manufacturer's instructions. Equal amounts of protein lysate were used for western blot analyses. Specific signals were visualized with NBT/BCIP.
Cell viability assay using MTT
Cells were seeded into 96-well plates, treated with different concentrations of cisplatin for 24 hours and then assessed using 3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide. Cell viability was determined by measuring the optical absorbance of cells at 570 nm wavelength and normalizing the values to the corresponding controls.
Analysis of apoptosis by flow cytometry
Cells were harvested, washed with PBS and stained with the annexin-V/PI apoptosis kit according to manufacturer's instructions. Analysis of apoptotic cells was performed using a FACScan flow cytometer, and the data were analyzed using cell fit software.
Immunofluorescence
Cells were trypsinized and plated onto chamber slides for 12 hours. After fixation in acetone, the slides were blocked with BSA, incubated with p21 antibody and rhodamine-conjugated secondary antibody and counterstained with DAPI. Slides were then observed on a confocal laser-scanning microscope (Olympus IX81, Japan). P21 staining was categorized as negative, nuclear or cytoplasmic according to a standard described before [
32]. Briefly, negative was defined as undetectable cytoplasmic or nuclear staining. Nuclear p21 was defined as the fraction of tumor cells with positive nuclear staining greater than or equal to that of positive cytoplasmic staining. Cytoplasmic p21 was defined as the fraction of cytoplasmic staining greater than that of nuclear staining. Five randomized fields were counted in order to calculate the percentage of cells that stained with anti-p21 antibody.
Immunohistochemistry
Paraffin-embedded tissue sections were deparaffinized. After antigen retrieval, slides were incubated with 3% H
2O
2 to inhibit endogenous peroxidase. Slides were then blocked with 5% normal serum and incubated with anti-p21 antibody. For qualitative identification of specific antibody staining, the DAKO Envision+ system was used, according to manufacturer's instructions, as described below. After detection with chromogen diaminobenzidine, sections were counterstained with hematoxylin and mounted. For the negative control, all incubation steps were identical except that PBS was used instead of primary antibody. The immunoreactivity of p21 was categorized as negative, nuclear or cytoplasmic according to the standard decribed before [
32].
Image analysis
Western blot images were captured and quantified using the ChemiImager 5500 system from the Alpha Innotech Corporation (San Leandro, CA, USA).
Statistical analysis
All experiments were repeated three times. The relationship between patients' clinical characteristics and results of p21 immunohistochemistry was assessed using the chi-squared (χ2) test. Results expressed as mean ± SD were analyzed using the Student t test. Differences were considered significant when p < 0.05. Data were analyzed using SPSS software version 13.0 (SPSS Inc., Chicago, IL).
Discussion
It is well known that p21
(Waf1/Cip1) in the nucleus functions as a tumor suppressor by binding to cyclin/CDK complexes and proliferating cell nuclear antigen [
15,
16]. However, recent studies [
31,
32] have revealed that p21 can be a paradoxical tumor promoting factor and has been associated with poor cancer prognosis due to its accumulation in the cytoplasm. Recently, Koster has reported cytoplasmic p21 expression levels determined cisplatin resistance in testicular cancer [
37]. This study was aimed at exploring the relationship between cytoplasmic p21 and cisplatin resistance in ovarian cancer, and investigating whether regulation of cytoplasmic p21 could alter the response to the therapy.
The cisplatin-resistant cell line C13* used in this study was established after in vitro challenge of OV2008 cells, a line derived from one ovarian carcinoma patient without prior chemotherapy, with cisplatin [
38,
39]. Since C13* and OV2008 cells shared the same genetic background and had minimum variation, they were superior to other cell lines and were chosen as optimal models for in vitro investigation of drug resistance. After confirming the expression levels and cellular distribution of p21 in the paired cell lines, we sought to determine whether it was influenced by cisplatin treatment and whether it was involved in cisplatin resistance. Long-term exposure to low-dose cisplatin resulted in p21 cytoplasmic translocation in cisplatin-sensitive cells, while the same treatment nearly had no effect on p21 cellular distribution in cisplatin-resistant cells. These results suggested p21's accumulation in cytoplasm was probably a protective response of sensitive cells to the drug, which helped them to escape from being killed by chemotherapeutics. Given that the endogenous p21 protein in resistant cells was mainly located in the cytoplasm, loss-of-function assay of cytoplasmic p21 was directly taken by RNA interference technology. Our results indicated knockdown of cytoplasmic p21 through p21 siRNA notably enhanced drug response in C13* cells. However, p21 transfection might not be the suitable strategy to illustrate the influence of increased cytocymic p21 because endogenous p21 in OV2008 was mainly restricted in the nuclear; The exogenous p21 transfected into OV2008 would probably mainly increase the nuclear amount instead of cytoplasmic amount. Thus, we increased the amount of cytoplasmic p21 by promoting its translocation from nuclear to cytoplasm in sensitive cells.
Zhou et al. reported that activation of phosphatidylinositol 3-kinase (PI3K)/Akt signaling can stimulate cytoplasmic accumulation of p21 [
29]. There are three isoforms of Akt: Akt1, Akt2 and Akt3. All three isoforms share a high degree of amino acid sequence identity, especially within the kinase domain [
40], and are activated by similar pathways in a PI3K-dependent manner. Overexpression of Akt2 has been found in 20% of human ovarian cancers [
41], and increased Akt2 kinase activity has been found in approximately 30% of ovarian cancers [
42]. Given these findings, we performed experiments where we altered the levels of cytoplasmic p21 by modulating Akt2 signaling. In OV2008 cells, a constitutively active AAkt2 plasmid was applied to activate PI3K/Akt signaling. Our results showed that the AAkt2 plasmid efficiently activated phosphorylation of Akt and, more importantly, significantly promoted the translocation of p21 into the cytoplasm. In the meantime, functional inhibition of cytoplasmic p21 was accomplished by short hairpin RNA silencing of Akt2 in C13* cells. Nevertheless, in fibroblasts and myoblasts, it has been suggested that the accumulation of p21 in the cytoplasm is stimulated by Akt1 [
43]. The differences in our results might be ascribed to the different cell types used in the studies, similarity among three Akt isoforms and the mutual activation of the different isoforms. Therefore, the role of different Akt isoforms in the cytoplasmic translocation of p21 requires further investigation.
Caspase 3 activation is considered to be a key cellular component of the terminal and irreversible phase of apoptotic death caused by DNA damaging agents. In our experiments using flow cytometry, we found that the levels of cleaved caspase 3 (17 kDa) were proportional to the rate of apoptosis. Previously, it was reported that p21 could bind to procaspase 3 [
31] and prevent its conversion to mature caspase 3, leading to the inhibition of apoptosis. In our experiments, it is likely that caspase 3 activation is inhibited by cytoplasmic p21; however, further studies are required to confirm this hypothesis.
By immunohistochemical analysis we found that a large proportion of tissues from treatment non-response tumors stained positive for cytoplasmic p21 (5 to 8) when compared to tissues from the treatment response group (2 to 25) (p = 0.027). We also showed that there was no significant difference in the level of cytoplasmic p21 staining between high stage and low stage ovarian tumors (p = 0.691), and between patients of different ages (p = 0.677). Furthermore, we compared cytoplasmic p21 levels in serous and non-serous tumors because the number of mucous and other types of tumors was limited. However, information on the prognosis and survival of the patients was not available; therefore, our investigation could not make a statistical analysis between cytoplasmic p21 and disease-free or survival rates.
Acknowledgements and Funding
This work was supported by the National Science Foundation of China (81101971, 30571950, 30901586, 30801340), Major Innovation Medicine program (2009ZX09103-738), "973" Program of China (No. 2009CB521808), Science Foundation of Guangdong Province (No. B2011295, No.S2011040006012), Shenzhen Scientific Program (No.201002006, No.20110422597) and Nanshan Scientific Program (No. 2010028). We thank Dr. Li Cao for reviewing and collecting clinical samples, and Dr. Qiling Ao for analyzing the immunohistochemistry results.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
XX and QM carried out the immunofluorescence staining, participated in functional assay and drafted the manuscript. XL, TJ, PC and HX participated in nucleus p21 siRNA sequence designing and silencing efficiency. KL, YF and DW performed cell viability assay, western blot and RT-PCR. YW, SL and ZH collected clinical data, and carried out flow cytometry and immunohistochemistry assay. RL and TZ performed statistical analysis. GX participated in the data interpretation and provided expertise in molecular biological techniques. SW was responsible for writing and revising the manuscript. JZ, DM and LM participated in the design and coordination of the study. All authors have read and approved the final manuscript.