Nuclear β-catenin accumulation is associated with increased expression of Nanog protein and predicts poor prognosis of non-small cell lung cancer

Primary tumor specimens from 309 NSCLC patients who underwent curative surgical resection were evaluated in this study. None of these patients had undergone chemotherapy or radiotherapy prior to the surgery. Of these patients, 203 were collected from 2001 to 2005 at Hunan Tumor Hospital and the rest of 96 were from 2002 to 2006 at the Affiliated Hospital of Guilin Medicine University. Patients in the study were examined and treated according to provincial guidelines. This study was approved by the Research Ethics Boards of the Southern Medical University, China.

Cores from formalin-fixed paraffin embedded tumor tissues of NSCLC were arrayed in triplicate onto a tissue microarray (TMA) and each TMA block consisted of up to 40 tissue cores, before the area of cores were chosen, the preliminary experiment of β-catenin were made to ensure the typical cores. TMAs were constructed using an automated tissue arrayer (Beecher Instruments Inc. USA) by Auragene Bioscience Corporation, China. Tissues core with a diameter of 3 mm was transferred to a recipient paraffin block (array margin of 15mm×20 mm). The block was sectioned at a series of 4-μm-thick slices. Once the TMA slices were made, they were heat 60°C for 1 h to aid cutting.

Relevant clinical pathologic features (Table 1) were obtained from the patients’ files and/or by telephone interviews with the patients or their relatives. Pathological stages were classified according to the TNM staging system. Histological grading and typing of the tumor were determined according to the World Health Organization tumor classification system.

Serial tumor sections and IHC staining were used to evaluate both β-catenin and Nanog proteins. β-catenin immunoactivity was examined using a mouse polyclonal antibody (Cell Signaling Technology, Danvers). Nanog protein was detected with a rabbit polyclonal antibody (Cell Signaling Technology, Danvers). The TMA blocks containing tissue sections (4 μm) were de-waxed in xylene, rehydrated through a graded series of ethanol solutions, rinsed in distilled water for 5 min, and then immersed in boiling citrate buffer in methanol for 1.5 min for antigen retrieval. After deparaffinization and rehydration, blocks were incubated in 3% hydrogen peroxide to block endogenous peroxidase activity. Then the sections were subject to immunohistochemistry staining using the primary antibodies overnight at 4°C in a humidity chamber. The avidin-biotin technique was applied using DAB for visualization and hematoxylin for nuclear counterstaining. Negative controls were prepared by omitting the primary antibody. Histological and IHC evaluation were independently performed by two pathologists without knowledge of the clinicopathological outcomes of the patients. We classified the IHC staining results into three categories according to subcellular localization of β-catenin, e.g. membranous, cytoplasmic and nuclear. Briefly, each slide was examined in its entirety under a light microscope, and an initial score was assigned which represented the estimated proportion of positive tumor cells (0: ≤5%; 1: 5~25%; 2: 25~75%; 3: ≥75%). The score 0 was defined as low and 1, 2, 3 as high. Slides with indeterminate evaluation were re-evaluated, and a consensus was reached.

For immunofluorescent staining, A549 [16] or H23 [17] cells grown on the surface of cover slides were serum-starved overnight followed by stimulation with 50 ng/ml EGF for 24 hours. Cells were fixed with 4% paraformaldehyde, rehydrated in PBS, and incubated with mouse anti-β-catenin (Santa Cruz Biotech) and rabbit anti-Nanog primary antibodies (Cell Signaling Technology, Danvers) at room temperature for 40 min. Subsequently, cells were incubated with Alexa Fluor 488-conjugated anti-rabbit antibody and Alexa Fluor 594-conjugated anti-mouse antibody (Molecular Probes; Invitrogen Corp.) for 40 min at room temperature. The nuclei were stained with DAPI. Slides were examined with a fluorescent confocal microscope (Olympus FV1000, Japan).

Cytoplasmic and nuclear proteins of A549 and H23 cells were extracted by Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime), then equal amount of protein were subjected to electrophoresis on a SDS-PAGE gel. The separated proteins were transferred to PVDF membranes (Millipore) and probed with appropriate primary antibodies. Protein bands were detected by enhanced chemiluminescence reagents (Pierce Biotechnology).

Both pLKO.1 lentiviral shRNA vector and control shRNA targeting against GFP were from Aldrich -Sigma. β-catenin targeting sequences are GCTTGGAATGAGACTGCTGAT, which was previously described in our earlier study [15]. The sense and antisense oligonucleotides were annealed and ligated into pLKO.1 lentiviral vector. The viruses were packaged in 293T cells according to standard protocols. Viral production and infection of target cells were previously described [18]. Infected cells were selected with 2 μg/mL puromycin.

Tumor specimen from 309 NSCLC patients consisting of 134 males and 175 female were included in this study. The median age was 52 years (range: 37–82 years). Of these patients, 167 (32.9%) were diagnosed as adenocarcinoma and 142 (33.1%) diagnosed as squamous carcinoma. Clinical stages, histological grades, and studied proteins were correlated as shown in Table 1. The tumor tissues was classified in three histological grades: grade 1 (126 patients), grade 2 (112 patients) and grade 3 (71 patients). The expression of membranous and nuclear, but not cytoplasmic, β-catenin expression were correlated with the histological grade (p<0.05), as shown in Table 1. Nanog expression that was localized in the nucleus was also correlated with the histological grade (p <0.01, Table 2).

Representative images of β-catenin immunohistochemical staining in NSCLC tissues are shown in Figure 1A-C. Positive membranous, cytoplasmic, and nuclear expression of β-catenin was detected in 67.0% (207/309), 43.0% (133/309), and 45.0% (139/309) of the NSCLC tissues, respectively.

Nanog is critically involved in regulation of cancer stem cells in several types of tumors and has been reported to be transcriptionally regulated by β-catenin. Thus we further examined Nanog expression in NSCLC specimen. We found that 30.4% (94/309) tumor tissues displayed Nanog immunoactivity that was located in the nucleus in most cases (Figure 1D).

The last follow-up date is Sep. 29th, 2009, with a median follow-up time 52 months (range 7–69.5 months). The 1-, 3- and 5-year overall survival (OS) rates were 82.4%, 52.5%, 30.6%, respectively. The survival rates of 309 NSCLC patients according to status of β-catenin and Nanog were shown in Table 2. The survival rate of patients with nuclear β-catenin expression was significantly lower than that of patients without nuclear β-catenin expression (p <0.01, Figure 2C). Likewise, the survival rate of patients with cytoplasmic β-catenin expression was significantly lower than that of patients without cytoplasmic β-catenin expression (p <0.01, Figure 2B). However, there was no statistical difference in the survival rate between patients with or without membranous β-catenin (p=0.064, Figure 2A), but with a trend toward reduced survival in patients with loss of β-catenin expression. Patients without Nanog protein expression had a significantly better prognosis and longer survival. The overall survival rate was 12.3% (Nanog-positive) and 38.3% (Nanog-negative) for this group (p <0.01) (Figure 3).

Univariate analyses showed no significant association between OS and age (>52 yr vs. ≤52 yr), sex (female vs. male), histological subtype (adenocarcinoma vs. squamous carcinoma), T stage (T3-T4 vs. T1-T2), clinical stage (stage III vs. I-II) (Table 3). The expression levels of cytoplasmic and nuclear β-catenin and Nanog were an independent prognosticator for OS (Table 3). In a multivariate analysis incorporating all clinicopathologic variables and covariates as shown in Table 3, comparisons between genders ages, tumor grades, tumor T stages, tumor C stages, expression levels of Nanog protein (low vs. high), and intracellular β-catenin protein expression at the membrane, in cytoplasm and in the nucleus (low vs. high) were performed to identify independent prognostic factors.

We further analyzed correlation between expression status of Nanog and β-catenin. As mentioned above, positive Nanog protein staining was almost exclusively observed in nucleus and β-catenin expression was stratified according to its subcellular locations (Table 4). Seventy two of the 139 (51.8%) tumor tissues that stained positive for nuclear β-catenin also displayed Nanog immunoactivity. Among the 170 specimen without nuclear β-catenin expression, only 22 (12.94%) showed Nanog immunoreactions (Fig. 4C, p < 0.001). On the contrary, only 55 (26.6%) of the 207 specimen with membranous β-catenin expression showed Nanog immunoreactions (Table 4, p = 0.036). Likewise, cytoplasmic β-catenin expression is not correlated with Nanog protein (Table 4, p = 0.166) even though cytoplasmic β-catenin is associated with poor prognosis. However, expression of cytoplasmic and nuclear β-catenin is correlated, among the 133 specimen with cytoplasmic β-catenin expression, 105 (78.9%) showed nuclear β-catenin immunoreactions (Table 5, p < 0.001).

EGFR signal transduction pathway is critically involved in cell proliferation and survival of tumors of epithelial cell origin [19]. Increased expression of EGFR has been detected in 40%–80% of NSCLC and has been associated with advanced disease and poor prognosis [20]. Recently, we demonstrated that activation of EGFR increased cancer stem-like cell properties in nasopharyngeal carcinoma, and this effect of EGFR was mediated by PI3K/AKT/β-catenin signaling [15]. In the present study, high incident of nuclear β-catenin was detected in lung cancer specimens and was associated with poor survival. Thus we asked if EGFR signaling could induce β-catenin activation in NSCLC. For this purpose, serum-starved A549 (Figure 4A) or H23 (Figure 4B) cells were incubated without or with EGF and subjected to immunofluorescent staining. In the absence of EGF, β-catenin was located predominantly at the plasma membrane, with faint staining distributed in the cytoplasm. When cells were treated with EGF, β-catenin staining translocated to the nucleus in both A549 and H23 cells. These results suggested that nuclear accumulation of β-catenin can be induced not only by Wnt signaling, but also by EGF stimulation in NSCLC. In addition, nuclear Nanog expression was dramatically enhanced in response to EGFR activation. These results of in vitro experiments with established cell lines are consistent with our finding that nuclear β-catenin is significantly associated with Nanog expression in primary NSCLC specimens.

We investigated the requirement of β-catenin in the regulation of Nanog in A549 and H23. To this end, β-catenin was knocked down with lentivirus-mediated shRNAs (Figure 5A). Compared with cells infected with the control shRNA lentivirus, cells infected with β-catenin shRNA lentivirus exhibited reduced Nanog expression. Moreover, Nanog expression in the NSCLC cells with knockdown of β-catenin can not be obviously enhanced by adding EGF, but on the other hand, when the expression of β-catenin is increased by adding EGF, Nanog thereupon increased expression upon β-catenin (Figure 5B). It is worthy of note that nuclear β-catenin rather than cytoplasm β-catenin expression in the NSCLC cells can be enhanced by adding EGF, (Figure 5C).