ARTICLE pubs.acs.org/jpr
Evaluation of P38 MAPK Pathway as a Molecular Signature in Ulcerative Colitis Xinmei Zhao,† Bin Kang,‡,§ Chaolan Lu,† Siqi Liu,‡ Huanjing Wang,† Xiaoming Yang,† Ye Chen,† Bo Jiang,† Jun Zhang,‡,§ Youyong Lu,*,‡,§ and Fachao Zhi*,† †
Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou 510515, China ‡ Beijing Genomics Institute, Chinese Academy of Sciences, Beijing, 101318, China § Laboratory of Molecular Oncology, Peking University Cancer Hospital/Institute, Beijing 100142, China
bS Supporting Information ABSTRACT: Early diagnosis and treatment of ulcerative colitis (UC) is clinically challenging. To overcome this problem, we explored the interrelated multiplex signaling pathway to identify molecular signatures in UC by using integrated strategy in proteomics. Intestinal mucosa of 12 UC cases and 12 normal controls underwent comparative proteomic analysis. A total of 26 unique differential proteins were identified, including 12 up-regulated and 14 down-regulated in UC group. A differential protein cluster, consisting of 11 proteins involved in p38 mitogenactivated protein kinase (MAPK) pathway, was deduced and validated by Western blot. Furthermore, three proteins elicited from the protein cluster, phosphorylated p38, MAWBP and galectin-3, as a molecular signature, were analyzed by immunohistochemistry on 118 UC and normal samples. Increased expression of P-p38 and down-regulated MAWBP and/or galectin-3 were detected in UC compared to normal samples (p < 0.001). This signature correlated with disease progression of UC (p < 0.01), and classified UC risk with high sensitivity (94.83 ( 2.91%) and specificity (98.33 ( 1.65%). In addition, P38 MAPK pathway modulated the expression of the protein clusters in macrophage cell line as evidenced by the alteration with specific inhibitor SB203580. These results indicate that molecular signature of P38 MAPK pathway might be a potential biomarker for evaluating UC risk. KEYWORDS: ulcerative colitis, P38 MAPK pathway, proteomics, MAWBP, molecular signature
’ INTRODUCTION Ulcerative colitis (UC), a subcategory of inflammatory bowel disease (IBD), is an idiopathic, chronic, relapsing, debilitating, and nonspecific inflammatory condition.1 UC may result in severe complications, such as toxic megacolon, colon cancer, intestinal hemorrhage, and bowel perforation.2,3 Severe UC has high risk of rehospitalization and colectomy.4 UC in earlier stage is difficult to be discriminated from other intestinal inflammatory disease such as infective enteritis, intestinal tuberculosis, ischemic colitis, Crohn’s disease (CD) and radiation enterocolitis.5 There is convincing evidence that suggests that the trend of UC incidence is rising,68 but the pathogenesis of this disease is not fully understood. To reduce the mortality and improve the efficacy of treatment, it is imperative to explore the interrelated multiplex signaling pathway to discover UC-specific molecular signature whose function may be involved in disease progression and which may be crucial for the understanding of the pathogenesis of the disease. The molecular signature in turn may be helpful for early detection of the disease and may serve as a potential therapeutic target of UC. Considering the complexity of UC, investigations aimed to obtain a coherent picture of the disease at protein level by large-scale screening approach are needed. However, there are two obstacles. r 2011 American Chemical Society
First, it is generally accepted that most targeted therapeutics are directed at proteins and multiple pathways of signaling transduction are involved in UC, whose functions are not only dependent on their expression abundance but also their status of post-translational modification.9 Second, UC is difficult to be discriminated from other inflammatory intestinal diseases. Therefore, the qualified specimens are required based on strict pathological diagnosis, long period followup of the patient, as well as active inflammation on a specific site of UC. In the past few years, a few efforts have been made to better define biological profiles of UC. Hsieh et al. used 2-DE and MS to identify differentially expressed proteins between the UC and normal colon mucosa. Thirteen down-regulated and six upregulated proteins were identified in the UC-diseased colon mucosa, five proteins, vimentin, galectin-3, HSPD1, prohibitin and TRA1 were characterized using Western blot.10 Shkoda et al. applied proteome analysis (PA) to characterize the protein expression profile of CD and UC, the comparison between primary IEC from IBD (CD and UC) and colon cancer patients identified 21 proteins with at least 2-fold changes in their Received: September 21, 2010 Published: March 23, 2011 2216
dx.doi.org/10.1021/pr100969w | J. Proteome Res. 2011, 10, 2216–2225
Journal of Proteome Research expression level. However, the proteomic data were not verified by IHC and further analyzed as a molecular signature for UC risk.11 In this study, we proposed an integrated strategy of proteomics approaches to discover biomarkers and generated a molecular signature for UC pathogenesis. We isolated and characterized a number of potential biomarker for UC classification through proteomics and bioinformatics analysis. Furthermore, we demonstrated that grouped P-p38, MAWBP, and galectin-3 as a molecular signature for UC development and progression.
’ MATERIALS AND METHODS Tissue Specimens and Cell Line
The institutional ethics committee approved all protocols and all enrolled subjects gave their informed consent. Total of 142 samples were collected by means of endoscopy in Nanfang Hospital, Guangzhou, China. UC diagnosis was conducted by senior pathologists and physicians based on HE staining and Mayo-score.12 UC biopsy samples were obtained from inflamed areas in UC patient. Control samples were obtained from normal colon tissue of the subjects in regular physical examination. After carefully referring other proteomic documents regarding analysis of clinical samples, we selected approximate 20 cases as a proper size for 2D proteomics. Twelve pairs of the sample were used to identify the differential protein cluster. The detailed clinical data of these patients were provided in Supplemental Table 1 (Supporting Information). The tissue samples were rinsed in normal saline to remove contaminants and were immediately frozen in liquid nitrogen until it was subjected to proteomic analyses. In addition, 58 UC and 60 normal tissue samples were obtained and subjected to IHC assessment. The detailed clinical data of these patients were provided in Supplemental Tables 2 (Supporting Information). RAW 264.7 cells, a murine macrophage cell line, were obtained from ATCC. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing L-glutamine supplemented with 100 U/mL penicillin, 100 μg/mL streptomycin, and 10% (v/v) heat-inactivated fetal bovine serum (FBS) at 37 C in a humidified incubator with 5% CO2. Protein Extraction and Two-Dimensional Electrophoresis
The TCA-acetone precipitation method was adopted in the sample preparation of intestinal tissues. Briefly, 40 mg of sample, which collected from 34 biopsies in the same inflamed region in the same patient, was crushed by metal mortar merged in liquid nitrogen, and then precipitated with 10% TCA/acetone for 3 h. The precipitate was washed with precooled acetone. After removing acetone with vacuum evaporation, the pellet was dissolved in lysis buffer containing 20 Mm Tris-HCl, pH7.5, 8 M urea, 4% 3-(3-cholamidopropyl) dimethylammonio-1propanesulfonate (CHAPS), 0.5% Pharmalyte (pH 310NL), 10 mM dithiothreitol (DTT), 1 mM phenylmethylsulfonyl fluoride (PMSF), and 2 mM ethylenediamine tetraacetic acid (EDTA).The lysate was sonicated with a probe sonicator for 5 min followed by centrifugation at 40 000 g for 30 min. After quantitative measurements of protein concentration by Bradford method, the supernatant was stored at 80 C until use for 2-DE or Western blotting. To extract the proteins from the cultured cells, the harvested cells were washed with PBS buffer and subsequently dissolved with lysis buffer followed by sonication with a probe sonicator for 5 min. After centrifugation at 40 000 g for 30 min and quantification
ARTICLE
of protein concentrations, the supernatants were directly subjected to Western blot. The extracted proteins were separated by 2-DE according to the manufacturer’s instructions. Briefly, two hundred micrograms of lysate of each sample was loaded onto an IPG gel strip (pH 310, Nolinear, 18 cm, GE Healthcare, Piscataway, NJ). The strips were rehydrated for 4 h at 0 V and for 8 h at 50 V. The isoelectric focusing was carried out with a program of gradient voltage at 500, 1000, and 8000 V, each for 1 h and finally remained at 8000 V until the value of Vh reached 66 000 by the PROTEAN IEF cell system (Bio-Rad, Hercules, CA). The focused strip were reduced with 1% DTT and alkylated with 2.5% iodoacetamide (IAM) in the buffer containing 6 M urea, 50 mM Tris-HCl, pH 8.8, 30% glycerol, 2% SDS, and trace bromophenol blue. The treated strips were subjected to 12% SDS-PAGE in Ettan DALT II System (Amersham Biosciences, Sweden) for the secondary electrophoresis. The proteins on the 2-DE gels were visualized using silver staining. The gels were fixed in 50% methanol and 5% acetic acid in water for 30 min, then washed in water for 1 h and incubated in 0.02% sodium thiosulfate solution for 2 min. After washing twice in water for 10 s each, the gels were stained in 0.1% silver nitrate solution for 30 min. Color development was allowed for 15 min in 2% sodium carbonate/ 0.04% formaldehyde, and then stopped with EDTA (1.46%). The stained gels were washed three times in water for 3 min each. All the silver stained 2-DE gels were scanned by POWER-LOOK 2100XL (UMAX, Dallas, TX) with 500 dpi. Relative spot volumes were estimated with ImageMaster 2-D Platinum software, version 3.0 (GE Healthcare, Piscataway, NJ). The 2-DE gels obtained from the samples of UC tissue and normal tissue were compared to identify the spots with significant changes in relative spot volume (>3-fold). Protein Identification by MALDI TOF/TOF MS
After image analysis, the differential spots were manually excised from gels and subjected to in-gel digestion with trypsin (Promega, US) at 37 C overnight. The resulting peptides were spotted onto Anchorchip (Bruker, US), mixed with CHCA (4 mg/mL) for cocrystallization, followed by washing in 0.1% TFA buffer for 30 s. Peptide mass spectra were obtained using ULTRAFLEX TOF/TOF MS (Bruker, Fremont, CA), and protein identification was analyzed by MASCOT search (http://www. matrixscience.com) against NCBI human database. As for the search parameters, the number of allowed miscleavages was set to one, and carbamidomethylation was chosen as fixed as well as oxidation of methionine as variable modification. Ion tolerance for peptide was set to 100 ppm. Probability score in MASCOT search above 66 was defined as significant. The proteins identified with low mascot scores were confirmed by tandem mass spectrometry and replicate identifications based on the parallel gels as well as the sample enrichment. For the tandem mass analysis, the higher peaks of signal intensity matched to certain protein were selected. With the analysis of the b-ions and y-ions generated from the parent ion and MASCOT searching, the sequences of selected peptide were obtained. To evaluate the false positives of the MASCOT search results, the reverse database for human proteins was constructed using RSF program generated from Beijing Proteome Research Center. In this database, all human proteins had their amino acid sequences reversed but retained consistent orders for biamino acid neighbors to arginine or lysine. Western Blotting
The proteins extracted from the tissues and cells were resolved in 10 or 12% SDS-PAGE and electrotransferred onto PVDF 2217
dx.doi.org/10.1021/pr100969w |J. Proteome Res. 2011, 10, 2216–2225
Journal of Proteome Research
ARTICLE
Figure 1. Proteomic profiles of UC and the control, examples of up or down-regulated spots in UC. (A) All 30 differential protein spots were respectively labeled with arrowheads on 2D gels. (B) Typical up-regulated spots in UC, Cdc42, annexinA2, TNK2, and (C) typical down-regulated spots in UC, galectin-3, HMGCS2, MAWBP. The experimental conditions were described in Materials and Methods. (D) Typical MS/MS spectrum from a down-regulated protein in UC, MAWBP. The mass spectrum of PMF generated from the digested peptides of the 2-DE spot, the MS/MS spectrum derived from the parent ion at 1659.855, in which the amino acid sequence, GEPGGQTQAFDFYSR, was deduced based upon these b-ions and y-ions.
2218
dx.doi.org/10.1021/pr100969w |J. Proteome Res. 2011, 10, 2216–2225
Journal of Proteome Research
ARTICLE
Table 1. Differential Proteins Up-Regulated in UC theoretical
experimental
value
value p
accession category Signal transduction
description
no.
sequence
function
pI
MW
pI
MW
fold
valuea
score
coverage (%)
annotation
Cdc42
gi|5542168
5.5
21
6.6
25
3.6
0.002
101
50
GTPase
TNK2
gi|14250828
9.5
40
7.4
61
¥
0.004
67
25
Downstream effectors of
ANXA2
gi|119597993
6.0
39
6.5
35
¥
0.007
85
33
Phospholipase inhibitor
Cdc42
Inflammatory response
HSP90B1
gi|62088648
5.1
66
4.8
55
¥
0.016
100
22
ATPase
HSPA5
gi|386758
5.0
72
4.7
47
¥
0.037
129
24
ATP binding
PSMA1
gi|119588883
6.6
34
6.4
34
¥
0.143
74
39
and Oxidative stress
ATP-dependent proteolytis
SOD2
gi|134665
8.4
25
7.0
25
4.2
0.148
72
28
Superoxide dismutase
TXNDC5
gi|42794771
5.6
48
5.4
51
¥
0.009
146
41
Isomerase
LSP1
gi|37964
4.7
37
4.7
25
¥
0.035
68
23
Neutrophil activation and
P4HB
gi|20070125
4.8
57
4.8
62
5.9
0.044
327
66
Protein disulfide isomerase
CathepsinD
gi|4503143
6.1
45
5.4
33
¥
0.019
68
20
chemotaxis Aspartic-type endopeptidase Energy generation
ALDOA
gi|4557305
8.3
40
7.4
44
¥
0.020
102
41
Glycolytic enzyme
PCYT2
gi|4505651
6.4
44
4.1
39
¥
0.107
68
20
Phospholipids biosynthetic
TPI1
gi|999892
6.5
27
6.3
30
¥
0.042
130
51
Triose-phosphate
Vimentin
gi|55977767
5.1
54
5.0
48
¥
0.017
358
76
Cytoskeleton constituent
RPSA
gi|125969
4.8
33
5.0
43
¥
0.118
113
36
Translational elongation
process isomerase Cell adherence and mobility a
P value of multiple testing (30 tests) was corrected by the Bonferroni approach with 30 times of original p value derived from the two-sample t-test. P values < 0.05 are in italics.
membranes (350 mA for 1 h). All blots were scanned using Imagequant ECL system (GE, U.S.A.). The antibodies for P-p38, p38, galectin-3, actin and cathepsin D were obtained from Santa Cruz, MAWBP and Cdc42 antibodies were generated in our laboratory by the immunization of rabbit or mouse with the recombinant proteins expressed in Escherichia coli. Immunohistochemistry (IHC) Assessment
A total of 118 UC and normal tissues were subjected to IHC analysis. Briefly, multiple 4-μm tissue sections were cut with a microtome (Instrumedics). H&E-stained sections were used for histological verification. IHC staining was performed using EnVisionþ Kit (Dako, Denmark). The slices were separately incubated with primary antibodies against P-p38, galetin-3 and MAWBP overnight at 4 C. The working concentrations were separately set as 1:20, 1:200 and 1:1000 (antibody/buffer). In this study, 58 cases UC were classified into two groups based on Mayo Score,12 mild group included Mild, while severe group included Moderate and Severe. For quantitative analysis, the ratio of positively stained cells to all cells in 5 random areas at 200-fold magnification was recorded. Scoring of tissue slides was performed independently by two investigators; the percentage of positive cells and the intensity of staining were scored from 0 to 3: 0, less than 10% of cells stained; 1, 1050% of cells stained; 2, 5075% of cells stained; 3, more than 75% of cells stained as described previously. Cytokine Analysis
Macrophage cell line RAW264.7 cells were treated with lipopolysaccharide (LPS) and the specific chemical inhibitor of
P38 MAPK (SB203580, Calbiochem, San Diego, CA) for 12 and 1 h respectively in 6 well plates. Briefly, RAW264.7 cells were seeded in 6-well plates at a density of 5 105 cells/m:. The cells were allowed to attach to the bottom of the plates for 24 h. Following the pretreatment with SB203580 (10 μm) for 1 h, cells were treated with LPS (1 μg/mL) for 12 h. The supernatants were collected and were stored immediately at 70 C until the time of analysis. TNF-R, IL-1β, IL-10 and IL-6 ELISA kits (R&D Systems, Minneapolis, MN) were used to measure these cytokines according to the manufacturer’s instructions. Statistical Analysis
Statistical analysis was performed by two-sample t-test assessing the difference between two groups (UC and control) of observations if p-values
