Heterogeneous Nuclear Ribonucleoprotein A1 Is ... - ACS Publications

Aug 28, 2009 - One of these proteins, heterogeneous nuclear ribonucleoprotein A1 (hnRNP A1) was chosen for validation and functional analysis. The dat...
0 downloads 0 Views 2MB Size
Heterogeneous Nuclear Ribonucleoprotein A1 Is Identified as a Potential Biomarker for Colorectal Cancer Based on Differential Proteomics Technology Yan-Lei Ma, Jia-Yuan Peng, Peng Zhang, Long Huang, Wei-Jie Liu, Tong-Yi Shen, Hong-Qi Chen, Yu-Kun Zhou, Ming Zhang, Zhao-Xin Chu, and Huan-Long Qin* Department of Surgery, The Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University, 600 Yishan Road, Shanghai 200233, People’s Republic of China Received April 22, 2009

Colorectal cancer (CRC) is the third most common cancer worldwide and has poor prognosis. To identify the proteins involved in colorectal carcinogenesis, we employed 2-DE and MALDI-TOF/TOF-based proteomics approach to study the differentially expressed proteins in tumor and adjacent nontumor tissue samples. Samples from 10 colorectal patients were analyzed. Of the 7 significantly and consistently altered proteins identified, hnRNP A1 was one of the most significantly altered proteins and its overexpression was confirmed using RT-PCR and Western blot analyses. Immunohistochemical examination showed that the enhanced expression of hnRNP A1 was correlated with the increasing severity of colorectal tissue and the progression of the colorectal cancer, as well as UICC (International Union against Cancer) staging, histo-differentiation, recurrence and decreased survival. By developing a highly sensitive immunoassay, hnRNP A1 could be detected in human serum and was significantly elevated in CRC patients compared with healthy volunteers. We proposed that hnRNP A1 could be considered as a novel serum tumor marker for CRC that may have significance in the detection and in the management of patients with this diease. Knockdown of hnRNP A1 expression by RNA interference led to the significant suppression of the cell growth in colorectal cancer SW480 cells in vitro. These data suggested that hnRNP A1 may be a potential biomarker for early diagnosis, prognosis, and monitoring in the therapy of colorectal cancer. Further studies are needed to fully assess the potential clinical value of this biomarker candidate. Keywords: colorectal cancer • proteomics • biomarker • diagnosis • prognosis • therapy • heterogeneous nuclear ribonucleoprotein A1

Introduction Colorectal cancer (CRC) is the third most common type of cancer and the fourth most frequent cause of death due to cancer worldwide.1,2 Up to 90% of the patients can be cured by surgery if the CRC is detected at an early stage, but unfortunately, the disease is very often diagnosed only an advanced stage, and prognosis is accordingly poor. Therefore, the early diagnosis and prognosis are urgent for proper control of CRC.2-4 Identification of biomarkers for early detection, prognosis, and response to treatment is an important goal for cancer research by multiplex technologies.5 Proteomic approaches are promising tools for the discovery of new cancer biomarkers and prognostic therapeutic drug targets.6-8 Proteomics has burst onto the scientific scene with stunning rapidity over the past few years, perhaps benefitting a discipline that can enjoy the virtually instantaneous conversion of a genome sequence to a set of predicted proteins.9 Proteomics * To whom correspondence should be addressed. Department of Surgery, The Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University, 600 Yishan Road, Shanghai 200233, People’s Republic of China. Tel.: +86 21 64361349. Fax: +86 21 64368920. E-mail: [email protected]. 10.1021/pr900365e CCC: $40.75

 2009 American Chemical Society

analysis is currently considered to be a powerful tool for global evaluation of protein expression, and proteomics has been widely applied in analysis of diseases, especially in the fields of cancer research.10,11 Some factors cannot be detected either by measuring the amount of RNA or by studying nucleotide sequence variation.12 Because of the association between protein alterations and malignancy, analysis of the cancer proteome can be more beneficial than genomics and transcriptomics.13 Previous studies have also involved the preliminary application of proteomics in the identification of the biomarkers for CRC.14-18 These studies reported slightly different findings and conclusions that few proteins were found to vary in concert; and the discrepancies might be due to their regional variability or tissue heterogeneity, or technical problems such as varying ability of mass spectrometry to identify a particular protein.12,19 It has been widely accepted that a major challenge to cancer proteomics is the integration of biochemical, genetic, and proteomics data in the detection of biomarkers, so as to provide the impetus for the next level of clinical application.20,21 In this study, differentially expressed proteins, between paired cancer Journal of Proteome Research 2009, 8, 4525–4535 4525 Published on Web 08/28/2009

research articles

Ma et al.

Table 1. Clinical Features of All Human Tissue Samples clinical features

Normal tissues Mean age (range, years) Gender (male/female) Total Adenoma Mean age (range, years) Gender (Male/female) Pathological type tubular adenoma villous adenoma mixed adenoma Total CRC Mean age (range, years) Gender (male/female) Location Ascending colon Transverse colon Descending colon Sigmoid colon Rectum UICC staging I II III IV (Liver metastasis) Differentiation grading Well-differentiated Moderately differentiated Poorly differentiated Total

number (%)

67.03 ( 11.31 (53-72) 88/64 152 64.59 ( 12.70 (48-63) 16/30 8 (22.2%) 16 (44.5%) 12 (33.3%) 36 (100%) 67.03 ( 11.31 (53-72) 88/64 44 (28.9%) 7 (4.6%) 13 (8.6%) 38 (25%) 50 (32.9%) 14 68 50 20

(9.2) (44.7%) (32.9%) (13.2%)

10 (6.6%) 98 (64.5%) 44 (28.9%) 152 (100%)

tissues and corresponding normal tissues, were profiled from 10 CRC patients. One of these proteins, heterogeneous nuclear ribonucleoprotein A1 (hnRNP A1) was chosen for validation and functional analysis. The data resulting from the study were expected to lead to an improved understanding of the involvement of different biomarkers in the initiation and developmental findings into clinical applications.

Materials and Methods Sample Collection and Preparation. Fresh CRC and paired adjacent normal tissues were obtained from 10 patients suffering CRC who underwent surgical resections at the Department of Surgery, the Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University from 2006 to 2007. A total of 152 paraffin-embedded CRC and paired adjacent normal specimens, and 36 adenoma specimens were collected between the years 1999 and 2003. A summary of clinical information for these patients was shown in Table. 1. Preoperative blood was obtained by venipuncture from 92 patients who were operated for CRC. Fifty-eight blood samples of healthy individuals were donated on a voluntary basis in the community of Shanghai. Mean age with SD and gender distribution of the patients were listed in Table 2. Written informed consent of all patients and blood donors was documented. The project was approved by the Scientific and Ethical Committee of Shanghai Jiao Tong University, China. 2D Gel Electrophoresis. A total of 100 mg of tissue sample was grinded into powder in liquid nitrogen, homogenized in 1 mL of lysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 30 mM 4526

Journal of Proteome Research • Vol. 8, No. 10, 2009

Tris-HCl, protease inhibitor cocktail) on ice, and sonicated (10 × 10 s pulses) on ice. The homogenate was subjected to centrifugation (12 000 rpm) for 1 h at 4 °C. The protein was precipitated with cold acetone at -20 °C for 2 h and dissolved with rehydration buffer (8 M urea, 2 M thiourea, 4% CHAPS, 100 mM DTT, and 2% ampholyte). Protein concentrations were determined by the Bradford method (Bio-Rad, Hercules, CA). IPG strips (18 cm, pH 4-7, NL, Bio-Rad, Hercules, CA) were passively rehydrated using 400 µL for 12 h at 17 °C. IEF was performed on an IEF Cell (Bio-Rad, Hercules, CA). The strips were equilibrated in equilibration buffer (25 mM Tris-HCl, pH 8.8, 6 M urea, 20% glycerol, 2% SDS, and 130 mM DTT) for 15 min, followed by the same buffer containing 200 mM iodoacetamide instead of DTT for another 15 min.22,23 Twelve percent SDS-PAGE gels were used for 2D gel separation. The gels were stained using CBB R-350 (Merck, Germany) according to the supplier’s protocol. The protein spots were detected, quantified, and matched using PD-Quest 2D-analysis software (Bio-Rad, Hercules, CA). Each sample was run in triplicate. Image Analysis. The images were scanned with a Bio-Rad GS-800 scanner (400-750 nm) and the differentially expressed proteins were identified using the PD-Quest 2D-analysis software (Bio-Rad, Hercules, CA). Spot intensity was quantified automatically by calculation of spot volume after normalization of the image by taking the ratio of intensity of one spot to the total spots, and expressed as a fractional intensity. Those spots with 1.5-fold (t test, P < 0.05) or more changes in expressed as a fractional intensity and frequencies higher than 25% were selected for identification. In-Gel Tryptic Digestion and Protein Identification by Mass Spectrometry. In-gel tryptic digestion and protein identification by mass spectrometry were carried out as described in Sun’s study10 and in our published proteomic study.11 Immunohistochemistry. Sections were stained with the rabbit anti-human hnRNP A1 antibody (diluted 1:200, Abcam) using the DAB substrate solution (Dako Cytomation GmbH, Hamburg,Germany)accordingtothemanufacturer’sinstructions. Cancer cells with hnRNP A1 cytoplasmatic and/or nuclei immunoreactivity were considered positive cells. For each section, a minimum of 5 representative fields with wellpreserved cancer tissue was examined at ×200 magnification, and 200 cancer cells were counted for each field. An average for immunostaining intensity or percentage of positive cells was taken over these fields. The immunostaining was evaluated as described previously,12,24,25 with some slight modifications. Immunostaining intensity (A) was classified as lack of staining (0), mild staining (1), moderate staining (2), and strong staining (3); the proportion of staining-positive cells (B) was semiquantitatively divided into 4 grades, as 75% (4). The score for each section was measured as A × B, and the result was defined as negative (-, 0), weakly positive (+, 1-3), positive (++, 4-7), and strongly positive (+++, 8-12). ELISA for hnRNP A1 and CEA. For detection of hnRNP A1 in human serum, a sandwich ELISA was developed using streptavidin-coated 96-well microtiter plates, with reference to Roessler’s study;13 the method was performed as described in our previously study.11 CEA was measured by a commercially available essay (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer’s instructions. siRNA Synthesis. A double strand siRNA oligonucleotide targeting hnRNP A1 (sense, 5′-GCUCUUCAUUGGAGGGUUGdT-

research articles

Overexpression of hnRNP A1 in Colorectal Cancer a

Table 2. Clinical Features of Tested Serum Samples

gender

samples

total number

age (years)

female

male

CEA (ng/mL)

hnRNP A1 (ng/mL)

Healthy controls CRC (all stages) UICC I UICC II UICC III UICC IV

58 92 31 22 21 18

52.83 ( 12.34 65.32 ( 10.15 65.01 ( 11.32 64.12 ( 12.03 65.14 ( 9.81 67.01 ( 7.44

22 40 13 7 9 11

36 52 18 15 12 7

1.52 ( 1.32b 5.73 ( 10.08b 2.84 ( 1.21c 2.96 ( 1.53c 4.78 ( 2.14c 38.65 ( 4.76c

35.81 ( 22.82d 121.25 ( 71.78d 88.61 ( 39.42e 112.57 ( 76.25e 133.64 ( 78.69e 150.18 ( 92.76e

a CEA and hnRNP A1 levels in serum were determined by ELISA, and median value (mean ( SD) are given for the indicated groups. Mean age ( SD and gender distribution for the individual groups are indicated. b Wilcoxon two-sample test, P < 0.01. c One-way ANOVA analysis, P < 0.01, LSD-t test, P < 0.01 (UICC I vs UICC IV). d Wilcoxon two-sample test, P < 0.01. e One-way ANOVA analysis, P > 0.05.

dT-3′; antisense, 3′-dTdTCGAGUUGUAACCUCCCAAC-5′) with dTdT-overhangs at 3′ end was designed according to the published sequence of human hnRNP A1. This sequence, corresponding to the nucleotides 45-63 bp, had been used in a previous study,26 and proved to be specific and effective. Scrambled siRNA nucleotides were used as a negative control, and siRNA nucleotides targeting GAPDH were used as positive control for transfection evaluation. Cell Culture and Transfection. SW480 cells, a human colon cancer cell line, was maintained in Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco) containing 10% fetal calf serum (Gibco), penicillin (100 U/L) and streptomycin (10 mg/ L) at 37 °C in an atmosphere containing 5% CO2. The cells were grown on a 6-well plate in growth medium until they reached 70% confluence; transfections of hnRNP A1 siRNA were carried out with Lipofectamine 2000 (Invitrogen Co.) according to the manufacturer’s instructions. Semiquantitative RT-PCR. Total RNA was isolated using Trizol reagent (Invitrogen). cDNA was synthesized using the QuantiTect Reverse Transcription kit (Qiagen) according to manufacturer’s instructions. PCR was performed using human hnRNP A1 primers (sense 5′-GGAGAAGCCATTGTCTTCGGA3′; antisense 5′-GCATAGGATGTGCCAACAATCA-3′). The amplification parameters consisted of 27 cycles at 94 °C for 30 s, 56 °C for 30 s, and 72 °C for 30 s. The PCR products (5 µL) were analyzed by electrophoresis in 1% agarose gels and visualized by SYBR Gold (Molecular Probes, Eugene, OR) staining. Western Blotting. Tissue samples were ground in liquid nitrogen and lysed in RIPA lysis buffer (50 mM Tris-HCl (pH 7.4), 0.25% sodium deoxycholate, 150 mM NaCl, 1% NP-40, 1 mM EDTA, 1 mM NaF, 1 mM Na3V4, 1 mM PMSF). The extracted proteins were separated by SDS-PAGE and immunoblotted using primary antibody against hnRNP A1 (Abcam). The blots were visualized by enhanced chemiluminescence reagents (Amersham Pharmacia Biotech). β-Actin was used as an internal control. MTT Assay. Cell growth and viability were assessed using an MTT cell proliferation kit (Roche). The cells were seeded on 96-well microplates at 1.0 × 104 per well at 24, 48, and 72 h post-transfection with hnRNP A1 siRNA. The cells were subsequently incubated with 10 µL of MTT labeling reagent for 4 h, followed by addition of 100 µL of solubilization solution into each well. The plates were left in the dark room overnight and optical density (OD) was measured at 590 nm test wavelength and a 620 nm reference wavelength with an ELISA multiwell spectrophotometer (MDC, Sunnyrale, CA). Cell vitality index was calculated with the following formula: VI % (vitality index) ) OD treated wells/OD controlled wells × 100.

Stastistical Analysis. Statistical calculations were performed by SPSS for Windows release 11.0.0 statistical software (SPSS, Inc.). All quantitative data were recorded as mean ( SD. Comparisons between two groups were performed by Student’s t test or Wilcoxon two-sample test; comparisons among multiple groups were performed by one way ANOVA, Dunnett-t test or LSD-t test; comparisons of ordinal data between two groups were performed by rank sum test; relevance analysis of ordinal data was performed by Crosstabs χ2 test. Survival curves were generated according to the Kaplan-Meier method and the statistical analysis was performed by Log-rank test. Statistical significance was defined as P < 0.05.

Results Differenially Expressed Proteins between CRC and the Corresponding Normal Tissues. Two-dimensional electrophoresis was performed with individual-matched CRC and normal colorectal tissues from 10 patients (mean age 60.30 ( 6.72 years; range 40-70 years). Image analysis was performed by PDQuest 7.0 software, and displayed well-resolved and reproducible protein profiles for both CRC and the corresponding normal tissues. Coomassie staining for CRC and the corresponding normal tissue, respectively, with a matching rate of 88.7%, visualized averages of 609 ( 53 and 620 ( 44 spots (Figure 1A,B). A total of seven proteins were identified, and all the information was listed in Table 3. For some proteins, differences between the experimental MW/pI and the theoretical value occurred; this may be due to post-translational modificationssuchastruncationand/orproteinphosphorylation.26,27 Notably, hnRNP A1 was found to show one of the most significant differences in expression between cancer and normal tissues (P ) 0.012). It was up-regulated more than 13fold in CRC, when compared with the normal tissue, and MS/ MS analysis revealed 5 matched-peptides, with 34% sequence coverage and a MOWSE score 124. The mass spectra of hnRNP A1 was shown in Figure 2. Validation of hnRNP A1 by Semiquantitative RT-PCR and Western Blotting Analysis. To confirm the altered expression of hnRNP A1 in CRC, validation experiment was carried out to compare its transcripts by RT-PCR analysis. A significant difference mRNA level of hnRNP A1 was observed between CRC and normal tissues (cancer tissues, 0.13 ( 0.01; normal tissues, 0.85 ( 0.05; Student’s t test, P < 0.01) (Figure 3A,B). Further, Western blot analysis was performed with anti-hnRNP A1 antibody and overexpression of hnRNP A1 was observed in all the cancer tissues examined (cancer tissues, 0.12 ( 0.08; normal tissue, 0.89 ( 0.01; Student’s t test, P < 0.01) (Figure 3C,D). Together, our data demonstrate that hnRNP A1 is overexJournal of Proteome Research • Vol. 8, No. 10, 2009 4527

research articles

Ma et al.

Figure 1. Proteomic analysis of CRC tissues and corresponding normal tissues using 2-DE gels. Whole cell lysates (1 mg) from (A) CRC and (B) normal tissues (adjacent nontumor tissue samples) were separated on 2-DE gel and visualized by Coomassie blue staining.

pressed in CRC tissues at both mRNA and protein levels; this is consistent with the observation in the 2-DE analysis. To further investigate the expression of hnRNP A1 in different clinicopathological stages, namely, normal colorectal tissue, adenoma, CRC, and CRC with liver metastasis, the experiment was performed using semiquantitative RT-PCR and Western blotting. The expression of hnRNP A1 at both mRNA and protein levels among four groups including normal colorectal tissues, adenoma tissues, carcinoma tissues, and carcinoma with liver metastasis tissues revealed a remarkable increasing trend (One-way ANOVA, P < 0.05) (Figure 3A,B), paralleled with the increasing severity of colorectal tissue and the progression of CRC. Analysis of hnRNP A1 Expression by Immunohistochemistry. To further validate the elevation of hnRNP A1 in clinical samples and with an aim to determine its role in CRC carcinogenesis and prognosis, immunohistochemistry was performed by examination of the hnRNP A1 expression pattern in paraffin-embedded tissues with different clinicopathological stages (Figure 4A-D). Localization of hnRNP A1 to the cytoplasm and nulei of the cancer cells revealed that the cancer cells were responsible for the overexpression of hnRNP A1 and that this was not attributable to the inflammatory response in the surrounding stroma. As shown in Table 4, significant differences in staining intensity and positive cells were observed among normal colon, adenoma, carcinoma and carcinoma with liver metastasis specimens (rank sum test, P < 0.01). The semiquantitative scoring of immunoreactivity for normal colon, adenoma, carcinoma and carcinoma with liver metastasis specimens was 1.17 ( 0.78, 2.41 ( 1.32, 6.92 ( 2.35 and 9.87 ( 3.12, respectively, which also suggested remarkable differences of hnRNP A1 immunoreactivity among normal colorectal tissues, adenoma, carcinoma and carcinoma with liver metastasis tissues (One-way ANOVA, P < 0.05). In addition, the correlation of hnRNP A1 expression in CRC with clinicopathological indexes was evaluated, including with gender, age, smoking and drinking habits, differentiation, location, clinical stage, recurrence, and so forth. The data indicated no apparent relationship between staining patterns and gender, age, location and habits. However, the patients who had advanced clinical stage disease, recurrent lesion, or poor differentiation showed increased hnRNP A1 expression (Table 5). Since the latter three factors have great impact on the prognosis of patients, we propose that hnRNP A1 is a good potential predictor of tumor outcome. 4528

Journal of Proteome Research • Vol. 8, No. 10, 2009

To further confirm the potential prognostic role of hnRNP A1, survival analysis was performed. Of the 152 CRC cases for survival analysis, immunoreactivity of hnRNP A1 was weakly positive in 15 patients, moderately positive in 67 patients, and strongly positive in 70 patients. The survival curves were generated according to the Kaplan-Meier method, and survival analysis using log-rank test suggested that immunoreactivity of hnRNP A1 was more likely to present with poor outcome of patients with CRC (log-rank test, P < 0.05; Figure 4E). The 5-year survival rates were 69.84, 51.29 and 26.51% for weakly positive, positive and strongly positive staining samples, respectively. Detection of hnRNP A1 in Human Serum. After having established up-regulation of hnRNP A1 in colorectal tissue, we assessed the potential release of this protein into the periphery and therefore its potential value as serologic biomarker for the disease. For this purpose, a highly sensitive immunoassay for hnRNP A1 was established. The hnRNP A1 mean level of 92 cases of CRC was 121.25 ( 71.78 ng/mL. The hnRNP A1 mean level of 58 normal control cases was 35.80 ( 22.82 ng/mL. HnRNP A1 level in CRC group was significantly higher than that in normal control group (Wilcoxon two-sample test, P < 0.0001) (Figure 5A). In addition, we also determined the levels of the established tumor maker CEA using a commercial immunoassay. The CEA mean level of 92 cases of CRC was 5.73 ( 10.08 ng/mL. The CEA mean level of normal control cases was 1.52 ( 1.32 ng/mL. CEA level in CRC group was significantly higher than that in normal control group (Wilcoxon twosample test, P < 0.01) (Figure 5B). CEA serum levels were stagedependent, being much higher in UICC stage IV (38.65 ( 4.67 ng/mL) than stage I (2.84 ( 1.21 ng/mL) (Wilcoxon two-sample test, P < 0.001). HnRNP A1 serum levels for UICC stage I, stage II, stage III and stage IV were 88.61 ( 39.42, 112.57 ( 76.25, 133.64 ( 78.69 and 150.18 ( 92.76 ng/mL, respectively, which suggested remarkable difference of hnRNP A1 serum levels among the four stages (One-way ANOVA, P < 0.05). HnRNP serum levels were also stage-dependent. The receiver-operating characteristic curve (ROC curve) and the area under curve (AUC) of hnRNP A1 and CEA were calculated. Relationship between the specificity and sensitivity of hnRNP A1 and CEA for the detection of CRC was shown in the ROC curve graph, respectively (Figure 5C). As a result, the area under the curve of hnRNP A1 and CEA was 0.83 ( 0.04 and 0.81 ( 0.04, respectively, and significantly higher than that of null hypothesis (true area was 0.5, P < 0.01). Thus, diagnostic accuracy of both markers was in a comparable range. This also

research articles

P35749

Q9NY65 Q00403

P62736

4

5 6

7

a All protein spots were identified by MALDI-TOF/TOF. b For several proteins, a few isoforms were identified in the same individual. c The number of peaks which match to the tryspin peptides. d The number of peaks which do not match to the trypsin peptides. e Protein score (based on combined mass and mass/mass spectra) were from MALDI-TOF/TOF identification. The proteins had statistically significant protein score and best ion score were considered successfully identified. f Ratio of spot intensity between colorectal cancer versus normal tissues (C/N) was quantified using PDQuest 2-DE software; “+” represents the spot intensity in colorectal cancer is higher than in normal tissues, “-” represents the spot intensity in normal tissues is higher than in colorectal cancer. g The p-values paired t test. h The peptides identified with statistically significant ion score. Each spot corresponding to one certain protein had at least one of the shown peptides identified.

K.SYELPDGQVITIGNER.F 0.014 -1.55 ACTA2

42381

5.23

11

14

38

183

0.031 0.032

Q3KNS6 3

Tubulin alpha-8 chain Transcription initiation factor IIB aortic smooth muscle

TUBA8 GTF2B

50746 35324

5.01 8.67

6 1

9 9

19 3

46 27

-2.21 -2.00

K.LDAFLVLEQLR.C; K.SLEADLMQLQEDLAAAE R.A R.QLFHPEQLITGK.E R.KAVELDLVPGR.S