Background
Lung cancer is one of the major causes of cancer-related deaths both in men and women throughout the world. The 5 year survival rate of lung cancer is as low as 15 % [
1]. Detection of lung cancer at early stage allows the patients to receive appropriate treatment with more options and less invasive surgery, thus decreasing the mortality and suffering of patients [
2,
3]. However, the disease-associated symptoms imposed a challenging situation for early cancer diagnosis. Those symptoms can be hardly felt by the patients themselves until the disease is progressed to an advanced stage, which makes lung cancer diagnosed at a late stage when the cancer cells have metastasized [
2]. Although chest x-ray check and spiral CT scans have been used in routine screening of lung cancer, they are ineffective methods with insufficient accuracy due to the limited technology capabilities, observer variations and data interpretations. Thus, the development of reliable biomarkers for early detection of lung cancer is an urgent field since there is no currently approved screening test for lung cancer detection.
Blood proteins have always been attractive targets for biomarkers in many different types of diseases including cancer. The expression change of blood proteins is associated with the progression of diseases [
4,
5]. Autoantibodies are antigen-driven responses that can be viewed as reporters from the immune system. The identity of antigens suggested antigens sensed by the immune system might play roles in the pathophysiology of the disease process [
5,
6]. In cancer patients, most of the autoantigens targeted by host’s immune system were identified as proteins in cell transformation and cancer progression e.g. mutant p53, K-ras, insulin-growth factor II mRNA binding protein family (IMPs), c-myc, p90/CIP2A, p16, p21, cyclin A, B and E [
6,
7]. Thus the identification of novel tumor-associated antigens may provide insights into proteins whose aberrant expression can be sensed by the immune system, and these cancer-associated antigens or antibodies may also be used as biomarker candidates for the development of non-invasive cancer detection method.
CIP2A was initially identified as a cancer-associated autoantigen in gastric and liver cancer. But the function of CIP2A was largely unknown due to the lack of homology of CIP2A to any other proteins [
8]. Until 2007, CIP2A was characterized as a cognate interactor of protein phosphatase 2A (PP2A). Functional analysis of the CIP2A by genetic, biochemical and animal studies revealed that CIP2A can selectively inhibit the phosphatase activity of certain PP2A complexes, stabilizing one of the PP2A targets, c-myc without affecting other PP2A substrates like mouse double minute 2 oncoprotein (MDM2). Although following studies showed that the expression of CIP2A was up-regulated in a variety of cancer types ranging from solid tumor to hematological malignancies, research on the cellular functions of CIP2A under physiological conditions and in diseases states are still limited [
9‐
16] . Only few CIP2A substrates had been identified so far including two proteins, protein kinase B (PKB/AKT) and UNCEB5. Depletion of CIP2A via siRNA in hepatocellular carcinoma, breast cancer and esophageal cancer not only decreased cancer cell proliferation but also sensitized cancer cells to chemotherapeutical agents like bortezomib and tumor necrosis factor. One of the mechanisms was the contribution of CIP2A to enhance AKT-mediated cell survival through the direct regulation of AKT-associated PP2A phosphatase activity. In addition, the expression of CIP2A in endothelial cells promotes cell survival under ligand-dependent cell signaling [
17,
18]. The latest finding of CIP2A in controlling cell-cycle progression dramatically advanced our understanding of CIP2A in cancer progression. Two independent groups individually found CIP2A directly bind to cell-cycle regulators including polo-like kinase (PlK1) and never in mitosis gene A-related kinase 2 (NEK2) [
19,
20]. CIP2A is able to translocate into the nucleus in mitotic entry, where it binds to Plk1 and became deposited in spindle pole [
19]. In addition, CIP2A was also found in centrosome at G2/M stage. Both of these studies showed that the depletion of CIP2A resulted in the extended cell division [
20]. These results implied that CIP2A has a broader role in regulating cell proliferation as previously expected [
21]. Therefore, more functional roles of CIP2A in cancer progression need to be explored.
During the search of novel tumor-associated autoantigens in lung cancer patients, we found that CIP2A was overexpressed in lung cancer and elicited autoimmune response in lung cancer patients. The depletion and overexpression of CIP2A via genetic methods decreased and increased cancer cell proliferation, respectively. To define the function of CIP2A in cell proliferation, phospho-array containing key phosphorylation molecules in cancer was used. Our studies showed that CIP2A positively regulates the phosphorylation of AKT, p38 kinase and c-Jun activated kinase (JNK). The further analysis of the JNK pathway showed that CIP2A is a positive regulator of mitogen-activated kinase 4 (MKK4)-JNK signaling pathway in response to growth factors. Inhibition of JNK activation by JNK inhibitor partially abolished CIP2A-induced cell proliferation. Therefore, our study proposed a novel function of CIP2A in cancer progression.
Methods
Cell culture
Non-small cell lung cancer cell lines, A549, NCI-H838, NCI-H1299 and NCI-H460 and two small cell lung cancer cell lines, H69 and H146 were purchased from Amercian tissue collection center (ATCC). All of the cell lines, except A549, were cultured in RPMI-1640 medium (Invitrogen, Carlsbad, CA) containing 10 % fetal bovine serum (FBS) (Invitrogen) at 37 °C with 5 % CO
2. A549 is cultured with the same conditions as for other cell lines, except that it is cultured in F12K medium (Invitrogen). The immortalized cell line BEAS-2B was purchased from ATCC and was cultured in LHC-9 medium (Invitrogen). HBEC3 cell was generally provided by Dr. Jerry W. Shay (University of Texas Southwestern Medical Center). This cell line was cultured in Keratinocyte-SFM supplemented with 50 μg/mL bovine pituitary extract and 5 ng/mL epidermal growth factor (Gibco) at 37 °C on porcine gelatin-coated tissue dishes as previously described [
22].
Serum sample
In this study, sera from 207 patients with lung cancer and 82 normal human sera were obtained from the serum bank in the investigator’s (Zhang) laboratory at The University of Texas at El Paso (UTEP), U.S.A. Normal human sera were collected from adult individuals who had no obvious evidence of lung diseases during annual health examinations. All lung cancer patients were diagnosed according to the criteria described in a previous study [
23]. Informed consent was obtained from all the participants, including patients and normal human beings, by our clinical collaborators. This study was approved by the Ethic Committee of The University of Texas at El Paso: Institutional Review Board of The University of Texas at El Paso.
Purification of autoantibodies from sera
To purify CIP2A-specific autoantibody from human sera, A549 total cell lysates were separated on 12 % SDS-PAGE gel and then transferred onto nitrocellulose membrane. The membrane around 90 kDa was cut off from the whole membrane, and then further cut into small pieces. Those pieces of membrane were incubated with individual serum containing autoantibody against 90 kDa autoantigen for 2 h at RT. The membranes were subsequently washed three times with TBST (10 min per washing cycle). Autoantibodies bound on the membrane were eluted with 1 M glycine solution, pH 3.0, which was immediately neutralized with 1 M Tris–HCl, pH 8.0 [
24].
ELISA
Recombinant GST-CIP2A fusion protein was purified using GST glutathione beads. The protein purity was >95 % by SDS-PAGE. Proteins were diluted in PBS to a final concentration of 0.5 μg/mL for coating polystyrene 96-well microtiter plates (Dynatech Laboratories, Alexandria, VA). A volume of 200 μL of each human serum samples at 1:200 dilutions was added to the antigen-coated wells and incubated for 1.5 h at RT. Horseradish peroxidase-conjugated goat antihuman IgG (Caltag Laboratories, San Francisco, CA) at 1:5,000 dilution and the substrate 2,2’-azinobis (3-ethylbenzthiazoline-6-sulfonic acid) (Boehringer Mannheim GmbH, Mannheim, Germany) were used as detecting reagents. Each sample was tested in duplicate, and the average OD at 405 nm was used for data analysis. The cutoff value designating positive reaction was the mean optical density (OD) of 82 normal human sera plus 3 standard deviations (SD). The detailed protocol of ELISA was used as described.
Tumor tissue array, H&E staining and immunohistochemistry (IHC)
Formalin fixed and paraffin embedded lung cancer tissue arrays and normal human lung tissue were obtained from Cybrdi (Rockville, MD). The lung cancer tissue array contained 71 individual cases and 72 cores. Each of the cores represents one donor. The array contained normal human lung tissues from 23 individuals, spotted in triplicates. For both of these two tissue slides, each array spot was 1.5 mm in diameter and 5 μm in thickness. All the slides were stained with H&E staining and verified by pathologists. IHC was performed according to manufacturer’s protocol. Tumor tissue arrays were deparaffinized in Xylene, followed by rehydration in alcohol and then boiled in citrate buffer for antigen retrieval. Before antibody incubation, tissue array slides were incubated with hydrogen peroxide to block any endogenous peroxide activity and normal mouse antibody to reduce the nonspecific binding. The mouse monoclonal CIP2A antibody was added to the slides at a concentration of 1:50 and incubated at room temperature for 1 h, and goat anti-mouse HRP was used as secondary antibody. After washing, the peroxides reaction was developed with 3,3’-diaminobenzidine (DAB). At last, the slides were covered with cover slides mounted with mounting media.
Both of the slides were examined by two independent, blinded investigators. Five views were examined per slide, and 100 cells were observed per view at 400× magnification. IHC staining of CIP2A was scored by following a semi quantitative scale (− to +++), which evaluated the intensity in tumor areas, and the percentage of cells showing significantly higher immunostaining than control cells in normal lung tissues. Nuclear and/or cytoplasmic immunostaining in tumor cells was considered positive staining. The intensity of CIP2A staining was scored as 1 (mild), 2 (weak), and 3 (strong). Percentage scores were assigned as 0, 0 %; 1, 1–50 % and 2, 50–100 %. The scores of each tumor sample were multiplied to give a final score of 0–6, and the tumors were finally determined as lower expression, score 0–3 and high expression, score 4–6.
Western-blot analysis
Cells were plated in 6-well tissue culture plates at 80 % confluence and incubated overnight. Cell lysates were obtained from transduced cells using cold radioimmunoprecipitation assay buffer [20 mmol/L Tris–HCl (pH 8.0), 100 mmol/L NaCl, 10 % glycerol, 1 % NP40, 0.5 % sodium deoxycholate]. Twenty micrograms of protein were separated in 10 % SDS-PAGE gels and wet transferred to nitrocellulous membrane (GE Healthcare Life Sciences), then blocked for 1 h at room temperature in TBS-T [50 mmol/L Tris–HCl (pH 7.5), 150 mmol/L NaCl, 0.1 % Tween 20] buffer containing 5 % nonfat milk. Membranes were then incubated overnight at 4 °C or 1 h at room temperature with the respective primary antibodies: CIP2A (1:500), pJNK (1:1,000), pATF2 (1:1000), phospho-c-Jun (1:1000) and actin (1:1,000). Anti-mouse or anti-rabbit secondary antibody conjugated to horseradish peroxidase (Santa Cruz Biotechnology) was used to visualize the stained bands with an enhanced chemiluminescence visualization kit (Santa Cruz Biotechnology).
Production of lentivirus-containing CIP2A short hairpin RNA and lentivirus-containing full length CIP2A
Two CIP2A short hairpin RNAs (knock-down), ligated in pLKO.1 vector and pLOX-KIAA1524 (overexpression plasmid containing the full-length CIP2A) were obtained from Open Biosystems. The mature antisense sequences for the two shRNA we used are as following:
Lentivirus was produced by co-transfection of pLKO.1 control (ligated with scramble sequence) or other pLKO.1-derived vector, or pLOC-KIAA1524 with pMD2.G, and pCMV-VSVG into HEK293T packaging cell lines. The supernatants containing lentvirus of HEK293T were harvested at 36 and 72 h post transfection. Supernatant were pooled, centrifuged to remove cells and then filtered through 0.45 μm low protein binding filter. Cells were plated in monolayer at different densities and infected with lentivirus constructs using 8 ng/mL polybrene. The stable cell lines were selected in the presence of 1 μg/ml puromycin (knock down) or 10ug/ml blastidinin (overexpression) for 2 weeks.
Screen of phospho-kinase array
Human phospho-kinase array were purchased from R&D system. This array contained 43 antibodies, which were pre-spotted in duplicate on nitrocellulose membranes. Before incubating with cell lysates, membranes were incubated with blocking buffer for 1 h at room temperature. Cell lysates (400 μg protein) were mixed with a cocktail of biotinylated detection antibodies. The mixture was then incubated with the array overnight at 4 °C. After the incubation, membranes were washed three times with washing buffer. Streptavidin-horseradish peroxidase and chemiluminescent detection reagents were subsequently added. Chemiluminescence was detected in darkroom with the same protocol for Western blot. Densitometry was measured with NIH Image J software version 1.45 to quantify the average pixel density/spotted area after normalization to the background controls using plug-in function, protein array. Data are representation of two independent experiments.
Statistical analysis
For comparisons between two groups, Student t test was used. Chi-square analysis was used for categorical outcomes. Both of these tests were performed in PRISM software (GraphPad Software, La Jolla, CA). All tests of statistical significance were two-sided. P values less than 0.01 and 0.05 were considered to be statistically significant.
Discussion
The relationship between the immune system and tumorigenesis is not only intrinsically important, but also fundamental to the design of future cancer therapies. The fact that proteins involve in tumorigenesis and cancer progression can be sensed by host’s immune system promoted researchers to profile proteins with immunogenic ability during the onset and progression of diseases [
26]. Our lab has well-established platforms to identify this type of proteins in cancer patients by either cDNA expression library screening or proteomic approaches [
27]. A series of such proteins had been identified like p62/IGF2BP2, p90/CIP2A, Hsp60 and Hsp70 [
8,
28,
29]. In this study, we not only found that autoantibodies against CIP2A had very high frequency (25.7 vs 4.9 % in normal) in the sera from lung cancer patients but also the overexpression of CIP2A in more than 60 % lung cancer tissues. Of notice, CIP2A was previously shown as a developmentally regulated protein, whose expression was restricted to embryonic stages but was silenced in normal adult tissues [
8]. Therefore, both of these evidences supported that aberrant CIP2A expression at early stage of cancer would trigger immune response. Although CIP2A is not secreted into the extracellular space nor present as membrane-bound form, it is possible that immune response was initiated when nascent cancer cell was recognized by the immune system and lysed to release intracellular proteins, which induce autoantibody generation. The presence of autoantibody against CIP2A in sera from lung cancer patients also suggested that CIP2A was activated at early stage of lung cancer initiation. Furthermore, the expression of CIP2A in pre-malignant stage contributed to tumor formation [
25]. And studies from other research groups reported that CIP2A overexpression were found in more than 70 % of other types of cancer patients [
9‐
16] . Taken together, the temporal expression pattern makes CIP2A as an ideal candidate for early cancer diagnosis. However, CIP2A may not be used as an individual diagnostic factor or as biomarker for certain types of cancer cancer; instead it can serve as a valuable factor when combining with other clinically established biomarkers for specific cancer detection.
During our evaluation of CIP2A expression in different lung cancer cell lines, one interesting observation is the expression of CIP2A in immortalized cell line BEAS-2B but not in another immortalized cell line HBEC3. Although many studies have demonstrated the overexpression of CIP2A in cancer cell lines, few studies had examined the expression in immortalized cell line. The expression difference between the pre-malignant cell lines may be due to the different genetic background. The BEAS-2B was immortalized with chimera viral protein while HBEC3 cell line was immortalized by inactivating only an individual protein, p16. The viral chimera protein may activate the transcription CIP2A but not by p16-mediated pathway. This is also consistent with the study that the expression of CIP2A does not statistically correlated with the expression of p53. In addition, the expression of CIP2A in immortalized cell line is an alternative evidence that CIP2A can be present in pre-malignant tissue as reported previously that CIP2A was present in pre-malignant tissue in the mouse studies as well as the in the patients’ tissues of head and neck cancer [
25]. Taken together with the results from other groups, our data implied CIP2A might be a prerequisite to cellular transformation.
Although abundant evidences had shown that CIP2A is overexpressed in many types of cancer, the role of this protein in cancer progression is still unknown. In our study, we found that CIP2A overexpression is correlated with increased phosphorylation of JNK. JNK bears both of tumor-promoting and tumor-suppressive function depending on the types and genetic contexts of the tumor cells [
30]. Although growth factors like EGF can activate the tumor-promoting function and the environmental stress like UV can activate the cell death pathway, the molecules in making this decision is nevertheless not determined. Our finding that CIP2A enhanced the activity of JNK signaling pathway and lead to the increased cell proliferation suggested CIP2A can be a regulator in determining the substrate selection and prime the cell fate for proliferation rather than apoptosis or necrosis. However, one challenging problem was to fish out the proteins targeted by CIP2A. The rationale behind this was based on the PP2A inhibitive ability of CIP2A. We have searched literature about the PP2A targets of JNK signaling pathways, it involves more than ten proteins upstream of MKK4/MKK7 with each has multiple phosphorylation events. We therefore failed to find the CIP2A targets in JNK signaling pathway due to the technique difficulties. Interestingly, another two identified PP2A inhibitors, I2PP2A and small t antigen have similar roles to CIP2A in promoting JNK phosphorylation and AP-1 transcriptional activity [
31]. However, the exact mechanism of how these PP2A inhibitors, including CIP2A, are able to modulate the JNK phosphorylation signaling in a direct or an indirect way is still unknown. A systematic investigation of the global phosphorylation change might be employed to identify the CIP2A targets in JNK signaling pathway.
Our phospho-antibody array has found the phosphorylation change of several other signaling molecules like AKT, GSK3beta and p38. The role of CIP2A in promoting AKT phosphorylation in a stress-dependent and growth-factor-dependent way has been well established. CIP2A in promoting AKT-mediated signaling through the effect on AKT-associated PP2A phosphatase activity has been confirmed in hepatocellular carcinoma, breast cancer and lung cancer cell lines. For GSK3beta and p38, we actually did not find overexpression or knockdown of CIP2A has direct effect on the phosphorylation of these two proteins by Western blot. And our results that CIP2A does not affect the protein level of beta-catenin indirectly supported that CIP2A does not regulate the phosphorylation of GSK3beta. Although MKK4/MKK7 activates both JNK and p38, we here only showed CIP2A regulates JNK phosphorylation.
Our study of CIP2A-mediated cell proliferation partially through priming JNK signaling for cell proliferation rather than cell apoptosis expanded the understanding of oncogenic function of CIP2A in cancer progression. An important aspect in our study is that JNK is only partially contribute to CIP2A-mediated cell proliferation indicating CIP2A has multi-targets, which synergistically to promote cell proliferation. The previous findings that CIP2A physically interacted with c-myc to increase stability, and bind with Plk1 and NEK2 to extend cell division are the possible reasons that inhibition of JNK pathway only partially rescued CIP2A-associated cell proliferation [
19,
20,
25]. Moreover, our recent evidences that CIP2A regulated cell proliferation through AKT/PKB signaling pathway, and promoting cancer metabolism enriched the network being played by CIP2A [
32,
33]. However, how these CIP2A targeted molecules and signaling pathways are united together in cancer progression require extensive systematic studies.
Competing interests
The authors report no conflict of interest.
Authors’ contributions
BP and JY designed the experiments. BP, YR, YL, and XX performed the experiments. BP, YR, YL, and XX interpreted the results. BP and JY draft the manuscript. JY provides the administrative support and funding. All authors read and approved the final manuscript.