High Levels of Nucleolar Expression of Nucleolin Are Associated with Better Prognosis in Patients with Stage II Pancreatic Ductal Adenocarcinoma

Published OnlineFirst July 13, 2010; DOI: 10.1158/1078-0432.CCR-09-3411

Clinical Imaging, Diagnosis, Prognosis Cancer Research High Levels of Nucleolar Expression of Nucleolin Are Associated with Better Prognosis in Patients with Stage II Pancreatic Ductal Adenocarcinoma

Lan Peng1, John Liang1, Hua Wang2, Xianzhou Song1, Asif Rashid1, Henry F. Gomez3, Lynda J. Corley1, James L. Abbruzzese2, Jason B. Fleming3, Douglas B. Evans3,4, and Huamin Wang1

Abstract Purpose: Nucleolin is a major nucleolar protein that has been shown to be overexpressed in rapidly dividing cells and plays an essential role in cell proliferation and survival. However, the expression and significance of nucleolin in pancreatic ductal adenocarcinoma (PDA) have not been studied. Experimental Design: We used a tissue microarray consisting of 1.0-mm cores of tumor and paired nonneoplastic pancreatic tissue from 69 pancreaticoduodenectomy specimens with stage II PDA. Nucleo- lin expression was evaluated by immunohistochemistry and scored quantitatively by image analysis. Nu- cleolin expression was classified as nucleolin-high or nucleolin-low using the median nucleolin labeling index of 3.5% as cutoff. Staining results were correlated with clinicopathologic features and survival. Results: Both PDAs and PDA cell lines showed nucleolar staining for nucleolin. Nucleolin expression was higher in PDAs and PDA cell lines than in nonneoplastic ductal epithelial cells. Among the 69 stage II PDAs, 34 (49%) were nucleolin-high. The median overall survival was 65.2 ± 16.3 months for patients who had nucleolin-high PDAs compared with 19.5 ± 3.3 months for patients whose tumors were nucleolin-low (P = 0.03, log-rank method). No significant correlation between nucleolin expression and other clinicopathologic parameters was found. In multivariate analysis, nucleolin expression was a prognostic factor for overall survival in patients with stage II PDA independent of patient's age, gender, tumor size, differentiation, and lymph node status. Conclusions: Nucleolin was overexpressed in PDAs and PDA cell lines. A high level of nucleolar expression of nucleolin was an independent prognostic marker for better survival for patients with stage II PDAs. Clin Cancer Res; 16(14); 3734–42. ©2010 AACR.

Pancreatic ductal adenocarcinoma (PDA) is the fourth Previous studies have shown that multiple genes are leading cause of cancer death in the United States (1). On- frequently altered in PDAs, including the activation of ly 15% to 20% of patients diagnosed with PDA are surgi- the K-ras oncogene by point mutation (2) and inactiva- cally resectable and most of the patients who undergo tion of the tumor suppressor genes p16 (3), SMAD4/ resection have stage II disease. The long-term survival rate DPC4 (4, 5), and p53 (6). More recently, it has been of patients who have curative resection is approximately shown that telomere shortening occurs at the early stage 20%, with a median survival of 18 to 24 months (1). during the development of PDA. Telomere shortening Despite pancreaticoduodenectomy, the disease commonly was detected in 91% of the pancreatic intraepithelial recurs, with the most common sites of recurrence being neoplasia 1A (PanIN 1A), the earliest putative precursor the liver, lung, peritoneal cavity, and pancreatic surgery lesion for PDAs (7). Telomere length is maintained bed; the prognosis is poor (1). by the human telomerase complex, which is composed of human telomerase RNA (hTR or TERC), and telomerase- associated protein 1 (hTEP1) and its catalytic subunit- Authors' Affiliations: Departments of 1Pathology, 2Gastrointestinal human telomerase reverse transcriptase (hTERT; refs. 8–10). Medical Oncology, and 3Surgical Oncology, The University of Texas M. Telomerase activity has been detected in 80% of surgi- D. Anderson Cancer Center, Houston, Texas; and 4Department of Surgery, The Medical College of Wisconsin, Milwaukee, Wisconsin cally resected PDAs and is increased in 75% to 95 % of pancreatic juice samples obtained from patients Note: L. Peng and J. Liang contributed equally to this study. with PDAs (11–13). These data suggest that telomere Corresponding Author: Huamin Wang, Department of Pathology, Unit 085, The University of Texas M. D. Anderson Cancer Center, 1515 shortening and telomerase play a role in the develop- Holcombe Boulevard, Houston, TX 77030. Phone: 713-563-1846; ment of PDA. Critical shortening in telomere length Fax: 713-563-1848; E-mail: [email protected]. may lead to progressive accumulation of chromosomal doi: 10.1158/1078-0432.CCR-09-3411 abnormalities and the progression of PanIN lesions to ©2010 American Association for Cancer Research. invasive PDA.

3734 Clin Cancer Res; 16(14) July 15, 2010

Nucleolin in Pancreatic Ductal Adenocarcinoma

IIA and 44 stage IIB) who had undergone pancreaticoduo- Translational Relevance denectomy at the University of Texas M. D. Anderson Cancer Center between 1990 and 2005 and who had Nucleolin is a major nucleolar protein that has been not received any form of preoperative chemotherapy shown to be overexpressed in highly proliferative cells and/or radiation therapy. Patients who received preopera- and is involved in many aspects of gene expression and tive chemotherapy and/or radiation and who died from regulation of telomerase. In this study, we showed that postoperative complications were excluded. Only two pa- nucleolin was overexpressed both in pancreatic ductal tients with stage I disease and three patients with stage IV adenocarcinomas (PDA) and PDA cell lines compared disease diagnosed by positive peritoneal washing on cytol- with nonneoplastic pancreatic ductal epithelial cells. A ogy at the time of surgery, but no patients with stage III high level of nucleolar expression of nucleolin was an disease, were identified in our study group. These five independent prognostic marker for better survival for cases were excluded from this study because the number patients with stage II PDAs. These findings are impor- of patients with stage I and IV disease were too small to tant because majority of the patients with surgically re- be representative. The median patient age at time of sur- sectable PDAs have stage II disease and little is known gery was 62 years (range, 40-80 years). Fifty-two patients about the prognostic markers in this group of patients. received postoperative chemotherapy and/or radiation therapy due to recurrence or distant metastasis. This study was approved by the Institutional Review Board of the The function and intracellular localization of the telo- University of Texas M. D. Anderson Cancer Center. merase complex are regulated in part by nucleolin, which is a multifunctional and mobile protein that can shuttle Tissue microarray construction among the nucleolus, nucleoplasm, cytoplasm, and cyto- To construct the tissue microarray used in this study, plasmic membrane (14, 15). Nucleolin is overexpressed in formalin-fixed, paraffin-embedded archival tissue blocks highly proliferative cells and is involved in many aspects and their matching H&E-stained slides were retrieved, re- of gene expression, including chromatin remodeling, DNA viewed, and screened for representative tumor regions recombination and replication, gene transcription, mRNA and nonneoplastic pancreatic parenchyma by a pathologist stability, etc. (15, 16). Cytosolic nucleolin has been shown (H. W.). For each patient, two cores of tumor and two cores to suppress the translation and induction of p53 after of paired nonneoplastic pancreatic tissue were sampled DNA damage (17) and to regulate the stability of BCL-2 from representative areas using a 1.0-mm punch. The mRNA in breast cancer cells (18). Nucleolin has also been tissue microarray was constructed with a tissue microar- shown to be expressed on the cell surface of endothelial rayer (Beecher Instruments) as described previously (24). cells and different types of cancer cells (19–21). Nucleolin expression on cell surface has been used as a marker for Immunohistochemical analysis for nucleolin angiogenic endothelial cells and functions as a receptor Immunohistochemical stain for nucleolin was done on μ for endostatin, which mediates the antiangiogenic and 4- m unstained sections from the tissue microarray blocks antitumor effects of endostatin (19, 21). More recently, using a mouse monoclonal antibody against nucleolin high levels of nucleolin expression have been shown to (4E2, Abcam Inc.). To retrieve the antigenicity, the tissue correlate with poor survival in patients with cutaneous mel- sections were treated at 100°C in a steamer containing anoma and pediatric intracranial ependymoma (22, 23). 10 mmol citrate buffer (pH 6.0) for 60 minutes. The sections Collectively, these studies suggest that nucleolin plays a role were then immersed in methanol containing 0.3% hydro- in human malignancies. However, nucleolin expression gen peroxidase for 20 minutes to block the endogenous and its significance in PDA have not been examined. Be- peroxidase activity and were incubated in 2.5% blocking cause majority of the patients with surgically resectable serum to reduce nonspecific binding. Sections were incu- PDAs have stage II disease and little is known about the bated for 90 minutes at 37°C with primary antinucleolin prognosis in this group of patients, we examined nucleolin antibody at a 1:2,000 dilution. Standard avidin-biotin expression in 69 pancreaticoduodenectomy specimens immunohistochemical analysis of the sections was done with stage II PDA and their paired nonneoplastic pancreatic according to the manufacturer's recommendations (Vector ductal epithelial cells. Using univariate and multivariate Laboratories). Diaminobenzidine was used as a chromo- analysis, we determined whether the expression level of gen, and hematoxylin was used for counterstaining. nucleolin was associated with survival and other clinicopathologic features in patients with stage II PDA. Quantitative measurement of nucleolin expression levels using computer-assisted image analysis The immunohistochemically stained slides of PDA tis- Materials and Methods sue microarrays were scanned at ×200 magnification with the Ariol 2.1 scanner and digital imaging instrument Case selection (Applied Imaging). The staining results were scored quan- We retrospectively reviewed the medical records and tis- titatively using the Ariol image analysis system (Applied sue specimens of 69 patients with stage II PDA (25 stage Imaging). To measure nucleolin expression in PDA cells

www.aacrjournals.org Clin Cancer Res; 16(14) July 15, 2010 3735

Peng et al.

ductal epithelial cells was measured as a ratio between the stained nucleolar area and the total nuclear area. The com- puterized images and the nucleolin labeling index were reviewed independently by two pathologists (L. P. and H.W.). Cases were categorized into two groups based on the median nucleolin labeling index in PDA specimens: nucleolin-high (nucleolin labeling index ≥3.5%) and nucleolin-low (nucleolin labeling index <3.5%).

Cell culture and Western blot analysis The pancreatic cancer cell lines Panc-1, AsPC-1, and Capan-1 were purchased from the American Type Culture Collection. The Panc-48, CFPAC-1, Panc-3, Panc-28, Capan-2, MIA PaCa-2, and Hs766T pancreatic cancer cells were generously provided by Dr. Paul Chiao (The Univer- sity of Texas M. D. Anderson Cancer Center). All cell lines were maintained either in DMEM or in RPMI-1640 medium supplemented with 10% fetal bovine serum in a humidified incubator containing 5% CO at 37°C. Fig. 1. Representative micrographs of tissue cores of nonneoplastic 2 and neoplastic pancreatic tissue with pancreatic ducts (A, B) and PDA The HPDE cell line, an immortalized normal pancreatic cells (C, D) marked and selected by hand-draw tool. The selected ductal cell line, was provided as a generous gift from nonneoplastic pancreatic ducts (B) and PDA cells (D) were used for Dr. Ming-Sound Tsao (Ontario Cancer Institute, Toronto, scoring the nucleolar expression levels of nucleolin. A to D, ON, Canada). Protein expression of nucleolin was ana- immunohistochemical staining for nucleolin; original scanning magnification, ×200. lyzed by 10% SDS-PAGE, which was electroblotted onto polyvinylidene difluoride membranes (Novex), blocked in 5% skim milk in 1× TBS, and probed with the antibo- and nonneoplastic pancreatic ductal epithelial cells, tumor dies against nucleolin or actin. Proteins were detected cells or nonneoplastic pancreatic ductal cells were selected using an enhanced chemiluminescence (ECL) kit (Amersham- and marked with a hand-draw tool. The stromal cells, pan- Pharmacia Biotech). creatic acinar cells, and islet cells were not included for the measurement of nucleolin expression levels. Figure 1 Nuclear extract preparation and shows representative micrographs of tissue cores and the coimmunoprecipitation assays selected PDA cells or nonneoplastic pancreatic ductal epi- MPanc-96 cells with 90% confluence were collected and thelial cells from the corresponding tissue cores for quan- washed with PBS buffer. The cell pellets were resuspended titation of nucleolin expression levels. The nucleolin in 800 μL of buffer A [10 mmol/L HEPES (pH 7.9), 10 labeling index in tumor cells or nonneoplastic pancreatic mmol/L KCl, 0.1 mmol/L EDTA, 1 mmol/L DTT, plus

Fig. 2. Representative micrographs showing no expression of nucleolin in benign pancreatic duct (A), poorly differentiated PDA with low nucleolin expression (B), moderately differentiated PDA with high nucleolin expression (C), and Panc-1 cells showing high levels of nucleolar expression of nucleolin (D). Original magnification, ×400. No cytoplasmic or membranous staining for nucleolin was detected.

3736 Clin Cancer Res; 16(14) July 15, 2010 Clinical Cancer Research

Nucleolin in Pancreatic Ductal Adenocarcinoma

cocktail of protease inhibitors] and lysed on ice for 15 difluoride membrane, and detected by Western blotting minutes. Then 50 μL of 10% NP-40 were added and the using anti-hTERT and antinucleolin antibodies. tubes were vortexed briefly and vigorously. The nuclear pellets were collected by centrifugation at 1,500 rpm for Statistical analysis 30 seconds. The nuclei were then lysed in 200 μL of buffer Categorical data were compared by χ2 analysis or C [20 mmol/L HEPES (pH 7.9), 0.4mol/L NaCl, 1 mmol/L Fisher's exact tests. Overall survival curves were con- EDTA, 1mmol/L DTT, plus cocktail of protease inhibitors] structed using the Kaplan-Meier method, and the log- in cold room with vigorous shaking for 15 minutes. Nu- rank test was used to evaluate the statistical significance clear extracts (300 μg) in 300 μLofproteinlysisbuffer of differences. The patients' follow-up information were incubated with 2 μg of antinucleolin or anti-hTERT through August of 2009 was extracted from the medical antibody (Santa Cruz Biotechnology Inc.) at 4°C for over- records and, if necessary, updated by review of the U.S. night with gentle shaking. Fifteen microliters of protein G Social Security Index. The recurrence information was up- for antinucleolin antibody or protein A for anti-hTERT an- dated every time a patient came to the clinic/institution tibody were added and incubated for 2 hours on ice to for a follow-up visit. Overall survival was calculated from pull down the immunocomplex. The precipitated proteins the time of surgery to the time of death from any cause were washed three times with protein lysis buffer, separat- or to the time of last follow-up, at which point the data ed by 10% SDS-PAGE gel, transferred to polyvinylidene were censored. The prognostic significance of clinical and

Fig. 3. A, mean nucleolar expression levels of nucleolin in nonneoplastic ductal epithelial cells and PDA cells. B, Western blots (WB) showing the expression of nucleolin in 10 different PDA cell lines and HPDE cells, an immortalized normal pancreatic duct cell line. β-Actin was used as a loading control. C, coimmunoprecipitation assays using nuclear extracts prepared from MPanc-96 and antinucleolin or anti-hTERT antibody followed by Western blot analysis as described in Materials and Methods.

www.aacrjournals.org Clin Cancer Res; 16(14) July 15, 2010 3737

Peng et al.

pathologic characteristics was determined using univari- pancreatic ductal epithelium, and nucleolin-low and ate Cox regression analysis. Cox proportional hazards nucleolin-high tumors are shown in Fig. 2A to C. Nucleolar models were fitted for multivariate analysis. After interac- staining for nucleolin was also present in all 10 PDA cell tions between the variables were examined, a backward lines included in our tissue microarray, and the nucleolin stepwise procedure was used to derive the best-fitting staining in Panc-1 cells is shown in Fig. 2D. The mean nu- model. Statistical analysis was done using Statistical cleolin labeling index in PDA cells was 5.8% ± 3.4% Package for Social Sciences software (for Windows 12.0, (median, 3.5%) in 69 samples. The nucleolar expression SPSS Inc.). We used a two-sided significance level of 0.05 of nucleolin was higher in PDAs compared with that in for all statistical analyses. paired samples of nonneoplastic pancreatic ducts (0.7% ± 0.5%, P < 0.0001, Fig. 3A). None of benign pancreatic ducts Results showed a nucleolin labeling index of ≥3.5% (the median nucleolin labeling index in PDAs). In contrast, 34 of 69 Nucleolar expression of nucleolin in benign (49%) PDAs showed a nucleolin labeling index of ≥3.5% pancreatic ductal epithelial cells and PDA cells (nucleolin-high). Similar results were obtained using anoth- Immunohistochemical stain for nucleolin showed nu- er mouse monoclonal antinucleolin antibody (clone 44F12, cleolar staining pattern in both PDA cells and benign pan- Novocastra; data not shown; ref. 25). To confirm our immu- creatic ductal epithelial cells. No cytoplasmic staining or nohistochemical staining results, we did Western blot anal- membranous staining was detected in either PDA cells or ysis to measure the expression levels of nucleolin in 10 benign pancreatic ductal epithelial cells. Representative different PDA cell line and HPDE cells, an immortalized micrographs showing nucleolin staining of nonneoplastic normal pancreatic ductal cell line. Consistent with our

Table 1. Clinicopathologic features and correlation of nucleolin expression in stage II pancreatic ductal adenocarcinomas

Characteristics No. Nucleolin-low (n = 35) Nucleolin-high (n = 34) P

Age (y) <60 25 10 15 0.29 60-70 27 14 13 >70 17 11 6 Gender Female 22 12 10 0.80 Male 47 23 24 Tumor size (cm) ≤2.0 15 6 9 0.39 >2.0 54 29 25 Tumor differentiation Well 10 5 5 0.20 Moderate 41 24 17 Poor 18 6 12 Margin status Negative 55 28 27 0.59 Positive 14 7 7 Lymph node status (stage) Negative (IIA) 25 13 12 1.00 Positive (IIB) 44 22 22 Postoperative chemotherapy No 14 8 6 0.77 Yes 55 27 28 Postoperative radiotherapy No 18 9 9 1.00 Yes 51 26 25 Recurrence or distant metastasis No 22 11 11 0.92 Local recurrence 15 7 8 Distant metastasis 32 17 15

3738 Clin Cancer Res; 16(14) July 15, 2010 Clinical Cancer Research

Nucleolin in Pancreatic Ductal Adenocarcinoma

Fig. 4. Kaplan-Meier curves of the overall survival and recurrence-free survival in patients with resected stage II PDAs. Patients whose tumors were nucleolin-high had a longer median overall survival (65.2 ± 16.3 months) than patients whose tumors were nucleolin-low (19.5 ± 3.3 months, log-rank method, P = 0.025). The difference in recurrence-free survival was not statistically significant.

immunohistochemical staining results, the expression levels extracts prepared from MPanc-96 cells and antinucleolin of nucleolin was significant higher in all 10 PDA cell lines or anti-hTERT antibodies. We found that antinucleolin compared with HPDE cells (Fig. 3B). antibody could coprecipitate endogenous hTERT and anti-hTERT antibody could coprecipitate endogenous nu- Nucleolin interacted with hTERT in MPanc-96 cells cleolin (Fig. 3C). These data showed that nucleolin inter- To examine the function of nucleolin in PDA cells, we acted with nuclear hTERT in MPanc-96 cells and may carried out coimmunoprecipitation assay using the nuclear regulate hTERT activity in pancreatic cancer cells.

www.aacrjournals.org Clin Cancer Res; 16(14) July 15, 2010 3739

Peng et al.

Correlation of clinicopathologic features with ranging from 40 to 80 years of age. According to the WHO nucleolin expression classification, 10 (15%) tumors were well-differentiated, Table 1 summarizes the clinicopathologic characteristics 41 (59%) moderately differentiated, and 18 (26%) poorly of the study population. There were 47 men and 22 women, differentiated adenocarcinomas. Twenty-five (36%)

Table 2. Univariate and multivariate analysis of overall survival in patients with stage II pancreatic ductal adenocarcinoma

No. HR (95% CI) P

Univariate analysis Age (y) <60 25 1.00 60-70 27 1.60 (0.73-3.50) 0.24 >70 17 2.09 (0.88-4.95) 0.09 Gender Female 22 1.00 Male 47 0.60 (0.31-1.15) 0.12 Tumor size ≤2.0 cm 15 1.00 >2.0 cm 54 1.95 (0.81-4.68) 0.13 Tumor differentiation Well 10 1.00 Moderate 41 2.67 (0.92-7.74) 0.07 Poor 18 2.30 (0.72-7.36) 0.16 Margin status Negative 55 1.00 Positive 14 1.29 (0.58-2.85) 0.53 Lymph node Negative 25 1.00 Positive 44 1.68 (0.83-3.40) 0.14 Nucleolin expression Low 35 1.00 High 34 0.48 (0.25-0.93) 0.03 Multivariate analysis Age (years) <60 25 1.00 60-70 27 1.34 (0.55-3.24) 0.52 >70 17 1.35 (0.42-4.29) 0.61 Gender Female 22 1.00 Male 47 0.58 (0.29-1.13) 0.11 Tumor differentiation Well 10 1.00 Moderate 41 3.16 (1.08-9.31) 0.04 Poor 18 3.56 (1.05-12.09) 0.04 Tumor size ≤2.0 cm 15 1.00 >2.0 cm 54 1.76 (0.73-4.25) 0.21 Lymph node Negative 25 1.00 Positive 44 1.33 (0.61-2.93) 0.47 Nucleolin Expression Low 35 1.00 High 34 0.39 (0.20-0.79) 0.01

Abbreviations: HR, hazard ratio; 96% CI, 95% confidence interval.

3740 Clin Cancer Res; 16(14) July 15, 2010 Clinical Cancer Research

Nucleolin in Pancreatic Ductal Adenocarcinoma

patients had stage IIA tumors and 44 (64%) had stage ductal epithelial cells. Given the fact that nucleolin plays IIB tumors. Fifty-five (80%) patients had R0 resection a critical role in the nucleolar localization and the func- (defined as all margins negative microscopically), 14 tion of hTERT (14, 15), it is possible that nucleolin co- (20%) had R1 resection, and no patient had a grossly pos- localizes with hTERT to the nucleolus in PDA cells and itive margin of resection. No significant associations were regulates the formation of active telomerase complex identified between nucleolin expression and patient's age, and biological functions of hTERT (28–30). Consistent gender, tumor size, differentiation, margin status, regional with this hypothesis, we showed that endogenous nu- lymph node metastasis, or recurrence. cleolin interacted with hTERT in the nuclei of PDA cells. Therefore, our findings support the notion that telomere Survival analysis shortening with the resultant chromosome instability, After pancreaticoduodenectomy, the median and upregulation of nucleolar expression of nucleolin, and mean follow-up times were 22.9 months and 38.0 activation of telomerase plays a role in the development months, respectively. No patient was lost to follow-up. of PDA. High levels of nucleolin expression correlated with bet- Using a mouse monoclonal anti-hTERT antibody (clone ter overall survival in patients with stage II PDA. The 44F12, Novocastra), recent studies indicated that nucleo- median overall survival was 65.2 ± 16.3 months for lar expression of hTERT is associated with worse survival patients who had nucleolin-high PDAs compared with and is a better predictor than established clinical indica- 19.5 ± 3.3 months for patients whose tumors were tors in patients undergoing curative hepatic resection for nucleolin-low (P = 0.03, log-rank method; Fig. 4A and colorectal metastases (31, 32), whereas others showed that Table2).Therewasnosignificantdifferenceinrecur- nucleolar hTERT expression is associated with better sur- rence-free survival between the patients whose tumors vival of patients with urothelial bladder cancer (33). How- were nucleolin-high and those with nucleolin-low tumors ever, this antibody (clone 44F12) has been shown to (P = 0.17; Fig. 4B). In multivariate analysis, nucleolin-high recognize nucleolin, but not hTERT (25). Using a mono- expression and tumor differentiation were prognostic clonal antibody against nucleolin, we found that a high factors for overall survival in patients with stage II PDA in- level of nucleolar expression of nucleolin in stage II dependent of patient's age, gender, tumor size, and lymph PDAs was associated with better overall survival and was node status (Table 2). an independent prognostic factor in patients with stage II PDAs. Similar results were obtained using the mouse Discussion monoclonal antibody (clone 44F12, data not shown). However, we did not observe significant correlation be- In this study, we found that nucleolar expression of tween nucleolar expression of nucleolin and other clinico- nucleolin was significantly higher in PDA tumor samples pathologic parameters in stage II PDAs. Better survival in and PDA cell lines than in nonneoplastic pancreatic duc- the nucleolin-high tumor group may be due to the fact tal epithelial cells or HPDE cells. A high level of nucleo- that nucleolin increases telomerase activity, which may lar expression of nucleolin was associated with better protect against genetic instability and prevent evolution survival and was an independent prognostic factor in of more aggressive tumor clones leading to more aggres- stage II PDAs. Our results suggested that nucleolar expres- sive behavior and distant metastasis (33, 34). Consistent sion of nucleolin may play a role in the progression of with this hypothesis, it was recently shown that low ex- pancreatic cancer. pression levels of hTERT mRNA in PDAs are associated Nucleolin is a major nucleolar protein and has been with worse prognosis and poor overall survival when com- shown to be overexpressed in rapidly dividing cells and pared with the patients whose tumors express high levels cancer cell lines. Nucleolin plays an essential role in cell of hTERT mRNA (35). proliferation and survival (16, 26). Conditional knock- In conclusion, our results showed that nucleolar ex- out of nucleolin markedly inhibits cell proliferation in pression of nucleolin was markedly increased in PDAs DT40 avian B-cell lymphoma cells (27). In this study, compared with nonneoplastic pancreatic ductal epitheli- we found that nucleolin was expressed at high levels on- al cells. A high level of nucleolar expression of nucleolin ly in the nucleolus of PDA cells. The nonneoplastic panin stage II PDAs was associated with a better prognosis creatic ductal epithelial cells also showed nucleolar and was an independent prognostic marker for stage II expression of nucleolin, but at significantly lower levels PDAs. than PDA cells. Consistent with our immunohistochemical staining results, all 10 PDA cell lines examined in this study showed significantly higher levels of nucleolin Disclosure of Potential Conflicts of Interest expression compared with HPDE cells, an immortalized No potential conflicts of interest were disclosed. normal pancreatic ductal epithelial cell line. Although nucleolin has been reported previously to be expressed in the cytoplasm and cytoplasmic membrane, we did Acknowledgments not detect cytoplasmic or membranous expression of nucleolin either in PDA cells or in nonneoplastic pancreatic We thank Dr. Jinsong Liu for providing valuable comments.

www.aacrjournals.org Clin Cancer Res; 16(14) July 15, 2010 3741

Peng et al.

Grant Support The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. AACR-Pancreatic Cancer Action Network Career Development Award in Pancreatic Cancer Research and the Institutional Research Grant from The Received 12/31/2009; revised 05/05/2010; accepted 05/10/2010; University of Texas M. D. Anderson Cancer Center. published OnlineFirst 07/13/2010.

References 1. Wolff RA, Crane CH, Li D, Abbruzzesse JL, Evans DB. Neoplasms of 20. Destouches D, El Khoury D, Hamma-Kourbali Y, et al. Suppression the exocrine pancreas. 6th ed. Ontario: B. C. Decker, Inc.; 2006. of tumor growth and angiogenesis by a specific antagonist of the 2. Almoguera C, Shibata D, Forrester K, Martin J, Arnheim N, Perucho M. cell-surface expressed nucleolin. PLoS One 2008;3:e2518. Most human carcinomas of the exocrine pancreas contain mutant 21. Shi H, Huang Y, Zhou H, et al. Nucleolin is a receptor that mediates c-K-ras genes. Cell 1988;53:549–54. antiangiogenic and antitumor activity of endostatin. Blood 2007;110: 3. Caldas C, Hahn SA, da Costa LT, et al. Frequent somatic mutations 2899–906. and homozygous deletions of the p16 (MTS1) gene in pancreatic 22. Mourmouras V, Cevenini G, Cosci E, et al. Nucleolin protein expres- adenocarcinoma. Nat Genet 1994;8:27–32. sion in cutaneous melanocytic lesions. J Cutan Pathol 2009;36: 4. Hahn SA, Schutte M, Hoque AT, et al. DPC4, a candidate tumor sup- 637–46. pressor gene at human chromosome 18q21.1. Science 1996;271: 23. Ridley L, Rahman R, Brundler MA, et al. Multifactorial analysis of 350–3. predictors of outcome in pediatric intracranial ependymoma. 5. Wilentz RE, Iacobuzio-Donahue CA, Argani P, et al. Loss of expres- Neuro-oncology 2008;10:675–89. sion of Dpc4 in pancreatic intraepithelial neoplasia: evidence that 24. Wang H, Wang H, Zhang W, Fuller GN. Tissue microarrays: applica- DPC4 inactivation occurs late in neoplastic progression. Cancer tions in neuropathology research, diagnosis, and education. Brain Res 2000;60:2002–6. Pathol 2002;12:95–107. 6. Redston MS, Caldas C, Seymour AB, et al. p53 mutations in pancre- 25. Wu YL, Dudognon C, Nguyen E, et al. Immunodetection of human atic carcinoma and evidence of common involvement of homocopo- telomerase reverse-transcriptase (hTERT) re-appraised: nucleolin lymer tracts in DNA microdeletions. Cancer Res 1994;54:3025–33. and telomerase cross paths. J Cell Sci 2006;119:2797–806. 7. van Heek NT, Meeker AK, Kern SE, et al. Telomere shortening is 26. Gorczyca W, Smolewski P, Grabarek J, et al. Morphometry of nearly universal in pancreatic intraepithelial neoplasia. Am J Pathol nucleoli and expression of nucleolin analyzed by laser scanning cy- 2002;161:1541–7. tometry in mitogenically stimulated lymphocytes. Cytometry 2001; 8. Feng J, Funk WD, Wang SS, et al. The RNA component of human 45:206–13. telomerase. Science 1995;269:1236–41. 27. Storck S, Thiry M, Bouvet P. Conditional knockout of nucleolin in 9. Harrington L, Zhou W, McPhail T, et al. Human telomerase contains DT40 cells reveals the functional redundancy of its RNA-binding do- evolutionarily conserved catalytic and structural subunits. Genes Dev mains. Biol Cell 2009;101:153–67. 1997;11:3109–15. 28. Etheridge KT, Banik SS, Armbruster BN, et al. The nucleolar locali- 10. Nakayama J, Saito M, Nakamura H, Matsuura A, Ishikawa F. TLP1: zation domain of the catalytic subunit of human telomerase. J Biol a gene encoding a protein component of mammalian telomerase is a Chem 2002;277:24764–70. novel member of WD repeats family. Cell 1997;88:875–84. 29. Yang Y, Chen Y, Zhang C, Huang H, Weissman SM. Nucleolar local- 11. Hiyama E, Kodama T, Shinbara K, et al. Telomerase activity is ization of hTERT protein is associated with telomerase function. Exp detected in pancreatic cancer but not in benign tumors. Cancer Cell Res 2002;277:201–9. Res 1997;57:326–31. 30. Liu K, Hodes RJ, Weng N. Cutting edge: telomerase activation in 12. Ohuchida K, Mizumoto K, Yamada D, et al. Quantitative analysis of human T lymphocytes does not require increase in telomerase re- human telomerase reverse transcriptase in pancreatic cancer. Clin verse transcriptase (hTERT) protein but is associated with hTERT Cancer Res 2006;12:2066–9. phosphorylation and nuclear translocation. J Immunol 2001;166: 13. Suehara N, Mizumoto K, Tanaka M, et al. Telomerase activity in 4826–30. pancreatic juice differentiates ductal carcinoma from adenoma and 31. Domont J, Pawlik TM, Boige V, et al. Catalytic subunit of human pancreatitis. Clin Cancer Res 1997;3:2479–83. telomerase reverse transcriptase is an independent predictor of 14. Khurts S, Masutomi K, Delgermaa L, et al. Nucleolin interacts with survival in patients undergoing curative resection of hepatic colo- telomerase. J Biol Chem 2004;279:51508–15. rectal metastases: a multicenter analysis. J Clin Oncol 2005;23: 15. Storck S, Shukla M, Dimitrov S, Bouvet P. Functions of the histone 3086–93. chaperone nucleolin in diseases. Subcell Biochem 2007;41:125–44. 32. Smith DL, Soria JC, Morat L, et al. Human telomerase reverse 16. Srivastava M, Pollard HB. Molecular dissection of nucleolin's role transcriptase (hTERT) and Ki-67 are better predictors of survival in growth and cell proliferation: new insights. FASEB J 1999;13: than established clinical indicators in patients undergoing curative 1911–22. hepatic resection for colorectal metastases. Ann Surg Oncol 2004; 17. Takagi M, Absalon MJ, McLure KG, Kastan MB. Regulation of p53 11:45–51. translation and induction after DNA damage by ribosomal protein 33. Mavrommatis J, Mylona E, Gakiopoulou H, et al. Nuclear hTERT im- L26 and nucleolin. Cell 2005;123:49–63. munohistochemical expression is associated with survival of patients 18. Soundararajan S, Chen W, Spicer EK, Courtenay-Luck N, Fernandes DJ. with urothelial bladder cancer. Anticancer Res 2005;25:3109–16. The nucleolin targeting aptamer AS1411 destabilizes Bcl-2 mes- 34. Hackett JA, Greider CW. Balancing instability: dual roles for telomer- senger RNA in human breast cancer cells. Cancer Res 2008;68: ase and telomere dysfunction in tumorigenesis. Oncogene 2002;21: 2358–65. 619–26. 19. Christian S, Pilch J, Akerman ME, Porkka K, Laakkonen P, Ruoslahti E. 35. Grochola LF, Greither T, Taubert HW, et al. Prognostic relevance of Nucleolin expressed at the cell surface is a marker of endothelial cells hTERT mRNA expression in ductal adenocarcinoma of the pancreas. in angiogenic blood vessels. J Cell Biol 2003;163:871–8. Neoplasia 2008;10:973–6.

3742 Clin Cancer Res; 16(14) July 15, 2010 Clinical Cancer Research

High Levels of Nucleolar Expression of Nucleolin Are Associated with Better Prognosis in Patients with Stage II Pancreatic Ductal Adenocarcinoma

Lan Peng, John Liang, Hua Wang, et al.

Clin Cancer Res 2010;16:3734-3742. Published OnlineFirst July 13, 2010.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-09-3411

Cited articles This article cites 34 articles, 17 of which you can access for free at: http://clincancerres.aacrjournals.org/content/16/14/3734.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/16/14/3734.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/16/14/3734. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.