HCC-significant gene and protein collection.

HCC-significant genes and proteins were collected from five HCC related databases, including OncoDB.HCC (Last modified on Apr. 2008, http://oncodb.hcc.ibms.sinica.edu.tw/index.htm), HCC.net (Update time: Apr 18, 2010, http://www.megabionet.org/hcc/index.php), dbHCCvar (Last modified on Sep 4, 2012, http://genetmed.fudan.edu.cn/dbHCCvar/), EHCO (Encyclopedia of Hepatocellular Carcinoma genes Online, Last modified on Sep, 2012, http://ehco.iis.sinica.edu.tw), and Liverome (Last modified on Apr 14, 2011, http://liverome.kobic.re.kr/). For HCC-related molecular events, OncoDB.HCC is the first comprehensive HCC oncogenomic database and contains 577 HCC-related genes [13]. HCC.Net, which has compiled a network with thousands of HCC-related genes identified in different types and stages of HCC, contains 2,131 HCC-related genes [14]. dbHCCvar is an online database which contains 636 human genetic variations and 195 human genes which have been statistically tested to be associated with HCC [15]. EHCO is derived from thirteen gene sets related to HCC and gives 3833 HCC-significant genes [16]. Liverome is a curated database of liver cancer-related gene signatures [17]. There are 6024 gene signatures obtained mostly from published microarray and proteomic studies, and thoroughly curated by experts. In order to facilitate data analysis, the different ID types for HCC significant genes and proteins were converted to ID from UniProtKB-Swiss-Prot/TrEmbL. The detailed information on these HCC significant genes and proteins is described in Table S1.

Protein-protein interaction (PPI) data.

PPI data were imported from eight existing PPI databases including Human Annotated and Predicted Protein Interaction Database (HAPPI) [18], Reactome [19], Online Predicted Human Interaction Database (OPHID) [20], InAct [21], Human Protein Reference Database (HPRD) [22], Molecular interaction Database (MINT) [23], Database of Interacting Proteins (DIP) [24], and PDZBase [25]. The detailed information on these PPI databases is described in Table S2.

Network construction.

HCC significant proteins were used to construct HCC related network. The PPI data were obtained from eight existing PPI databases as mentioned above. Then, we applied Navigator software (Version 2.2.1) and Cytoscape (Version 2.8.1) to visualize the networks.

Defining network topological feature set.

For each node i in HCC related network, we defined five measures for assessing its topological property: (1) 'Degree' is defined as the number of links to node i; (2) 'Node Betweenness' is defined as the number of edges running through node i. (3) ‘Closeness’ is defined as the inverse of the farness which is the sum of node i distances to all other nodes. The Closeness centrality can be regarded as a measure of how long it will take to spread information from node i to all other nodes sequentially. Degree, node betweenness and closeness centralities can measure a protein’s topological importance in the network. The larger a protein’s degree/betweenness/closeness centrality is, the more important the protein is in the PPI network [28]. (4) K-core analysis is an iterative process in which the nodes are removed from the networks in order of least-connected [29]. The core of maximum order is defined as the main core or the highest k-core of the network. A k-core sub-network of the original network can be generated by recursively deleting vertices from the network whose degree is less than k. This results in a series of sub-networks that gradually reveal the globally central region of the original network. On this basis, 'K value' is used to measure the centrality of node i. (5) 'Edge Betweenness' is defined as the frequency of an edge that places on the shortest paths between all pairs of vertices in network [30]. The edges with highest betweenness values are most likely to lie between sub-graphs.

Ethics Statement.

The study was approved by the Research Ethics Committee of 302nd Hospital of PLA, Beijing, China. Written informed consent was obtained from all of the patients. All specimens were handled and made anonymous according to the ethical and legal standards.

Patients and Tissue Samples.

A total of 130 patients with primary HCC who underwent a curative liver resection at the 302nd Hospital of PLA, Beijing, China, were included in this study. One hundred and thirty self pairs of HCC and adjacent nonneoplastic liver tissues (as control tissues) obtained from these patients with HCC were retrieved from the tissue bank of the Department of Pathology in the 302nd Hospital of PLA. These patients were diagnosed as HCC between 2001 and 2006. None of the patients recruited in this study had chemotherapy or radiotherapy before the surgery. HCC diagnosis was based on World Health Organization (WHO) criteria. Tumor differentiation was defined according to the Edmondson grading system. Liver function was assessed using the Child-Pugh scoring system. Tumor staging was determined according to the sixth edition of the tumor-node-metastasis (TNM) classification of the International Union against Cancer. The clinicopathological features of 130 patients are summarized in Table 1.

The median follow-up period was 8.6 years. Postoperative surveillance included routine clinical and laboratory examinations every third month, computed tomography scans of the abdomen, and radiographs of the chest every third month. After 5 years, the examination interval was extended to 12 months.

Immunohistochemistry analysis.

GRB2, GAB1 and phosphorylated ERK1 (p-ERK1) expression were immunohistochemically evaluated in paraffin-embedded specimens of 130 pairs of HCC and adjacent nonneoplastic liver tissues. Surgical specimens were fixed in 10% formalin, embedded in paraffin, and sectioned at a 4 μm thickness. For heat-induced epitope retrieval, deparaffinized sections were soaked in 10 mM citrate buffer (pH 6.0) and treated at 95°C for 30 min using the microwave oven method. Immunohistochemical staining was performed using the avidin-biotin immunoperoxidase technique according to our previous studies [31-33]. The activity of endogenous peroxidase was blocked by incubation with 0.3% H₂O₂ in methanol for 15 min, and nonspecific immunoglobulin binding was blocked by incubation with 10% normal goat serum for 10 min. Sections were incubated at room temperature for 4 h with anti-GRB2 rabbit polyclonal antibody (#ab111031, Abcam, Cambridge, United Kingdom), or anti-GAB1 rabbit polyclonal antibody (#ab59362, Abcam, Cambridge, United Kingdom), or anti-phosphorylated ERK1 (phospho Y204) rabbit polyclonal antibody (#ab131438, Abcam, Cambridge, United Kingdom), and were then rinsed and incubated for 30 min with a biotinylated second antibody. After washing, the sections were incubated for 30 min with horseradish peroxidase-conjugated streptavidin, and were finally treated with 3,3’-diaminobenzidine tetrahydrochloride in 0.01% H₂O₂ for 10 min. The slides were counterstained with Meyer’s hematoxylin. The negative controls were processed in a similar manner with PBS instead of primary antibody. The positive GRB2, GAB1 and p-ERK1 expression confirmed by western blotting were used as positive controls for immunostaining.

Following a hematoxylin counterstaining, immunostaining was scored by two independent experienced pathologists, who were blinded to the clinicopathological parameters and clinical outcomes of the patients. The scores of the two pathologists were compared and any discrepant scores were trained through re-examining the stainings by both pathologists to achieve a consensus score. The numbers of positive-staining cells showing immunoreactivity in the cytoplasm for GAB1 and p-ERK1, and in the cell nucleus and cytoplasm for GRB2 in ten representative microscopic fields were counted and the percentage of positive cells was calculated. The percentage scoring of immunoreactive tumor cells was as follows: 0 (0%), 1 (1-10%), 2 (11-50%) and 3 (>50%). The staining intensity was visually scored and stratified as follows: 0 (negative), 1 (weak), 2 (moderate) and 3 (strong). A final immunoreactive score (IRS) was obtained for each case by multiplying the percentage and the intensity score. Therefore, tumors with a multiplied score exceeding median of total scores for GRB2, GAB1 and p-ERK1 were deemed to be high expression groups; all other scores were considered to be low expression groups.

Statistical analysis.

The software of SPSS version13.0 for Windows (SPSS Inc, IL, USA) and SAS 9.1 (SAS Institute, Cary, NC) was used for statistical analysis. The chi-squared test was used to show differences in categorical variables. Patient survival and the differences in patient survival were determined by the Kaplan-Meier method and the log-rank test, respectively. A Cox regression analysis (proportional hazard model) was performed for the multivariate analyses of prognostic factors. Differences were considered statistically significant when P was less than 0.05.

Upregulation of GRB2 and GAB1 proteins in HCC tissues.

The subcellular localization and the expression pattern of GRB2 and GAB1 proteins in 130 self pairs of HCC and adjacent nonneoplastic liver tissues were observed by the immunohistochemistry analysis. As shown in Figure 5A and 5D, GRB2 positive staining was localized in the cell nucleus and cytoplasm, while GAB1 positive staining was localized in the cytoplasm of tumor cells in HCC tissues. Compared with the adjacent nonneoplastic tissues, the expression levels of GRB2 (IRS for HCC vs. nonneoplastic liver: 6.32±1.50 vs. 2.67±0.32, P<0.001, Figure 5C) and GAB1 (IRS for HCC vs. nonneoplastic liver: 5.72±0.95 vs. 1.75±0.48, P<0.001, Figure 5F) proteins were all significantly increased in HCC tissues. Based on the scoring system used in the present study, 56 (43.08%) cases were both high expression of GRB2 and GAB1, 54 (41.54%) cases were both low expression of GRB2 and GAB1, 9 (6.92%) cases were GRB2-high and GAB1-low expression, and 11 (8.46%) cases were GRB2-low and GAB1-high expression. As determined by Spearman’s correlation, the GRB2 expression was significantly associated with the GAB1 expression (r=0.68, P=0.01).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Representative immunohistochemical images of GRB2 (A and B), GAB1 (D and E) and p-ERK1 (G and H) expression in HCC and adjacent non-neoplastic liver tissues (Original magnification×400).

Statistical analyses of IRS for GRB2 (C), GAB1 (F) and p-ERK1 (I) immunostainings in HCC and adjacent non-neoplastic liver tissues. GRB2 positive staining was localized in the cell nucleus and cytoplasm, while GAB1 positive staining was localized in the cytoplasm of tumor cells in HCC tissues. Compared with the adjacent nonneoplastic tissues, the expression levels of GRB2 (IRS for HCC vs. nonneoplastic liver: 6.32±1.50 vs. 2.67±0.32, P<0.001) and GAB1 (IRS for HCC vs. nonneoplastic liver: 5.72±0.95 vs. 1.75±0.48, P<0.001) proteins were all significantly increased in HCC tissues. More interestingly, in all four groups according to the combined expression of GRB2 and GAB1, GRB2-high/GAB1-high patients expressed the highest level of p-ERK1 protein (all P=0.01).

https://doi.org/10.1371/journal.pone.0085170.g005

In order to validate the hypothesis that GRB2 and GAB1 can induce the activation of the HGF/MAPK/ERK pathway, we detected the expression of p-ERK1 protein in 130 self pairs of HCC and adjacent nonneoplastic liver tissues. As shown in Figure 5G, p-ERK1 positive staining was localized in the cytoplasm of tumor cells in HCC tissues. Compared with the adjacent nonneoplastic tissues, the expression level of p-ERK1 (IRS for HCC vs. nonneoplastic liver: 4.04±0.67 vs. 1.20±0.33, P<0.001) protein was significantly increased in HCC tissues. More interestingly, in all four groups according to the combined expression of GRB2 and GAB1, GRB2-high/GAB1-high patients expressed the highest level of p-ERK1 protein (all P=0.01, Figure 5I).

Upregulation of GRB2 and GAB1 proteins associates with the aggressive tumor progression of HCC.

To evaluate whether GRB2 and GAB1 protein expression was associated with clinicopathological features of patients with HCC, we correlated IRS of GRB2 and GAB1 proteins with tumor stage, tumor grade, serum AFP level, presence of cirrhosis, underlying liver disease including alcohol abuse, viral hepatitis B and C, sex, and age (Table 1). As the results, we found that the expression levels of GRB2 protein in HCC tissues with the higher tumor stage (T3~4) and the positive serum AFP level were significantly lower than those with the lower tumor stage (T1~2, P=0.01, Table 1) and the negative serum AFP level (P=0.006, Table 1), respectively. In addition, the frequencies of aberrant GAB1 expression were higher in HCC tissues with higher tumor stage (T3~4) than those with lower tumor stage (P=0.01, Table 1). GAB1 overexpression was also observed more frequently in HCC tissues with high tumor grade than those with low grade (P=0.02, Table 1). More importantly, the combined GRB2 and GAB1 protein expression was significantly associated with serum AFP (P=0.01, Table 1), tumor stage (P=0.006, Table 1) and tumor grade (P=0.02, Table 1).

Upregulation of GRB2 and GAB1 proteins predicts the poor prognosis in patients with HCC.

Five-year disease-free survival was observed in 30 (23.08%) patients, whereas in 100 (76.92%) patients, disease recurred, and 88 (67.69%) even died during a 5-year follow-up period. We observed a trend that 5-year disease-free survival in the group with high GRB2 expression was significantly poorer than that in the group with low GRB2 expression (P=0.002, log-rank test; Figure 6A). Additionally, the Kaplan-Meier plot of 5-year overall survival curves stratified by GRB2 expression was shown in Figure 6B. A significant relationship was found between GRB2 expression and 5-year overall survival (P=0.006, log-rank test, Figure 6B). Similar with GRB2, the disease-free survival (Figure 6C, P=0.006) and overall survival (Figure 6D, P=0.008) of HCC patients with high GAB1 expression were both significantly shorter than those with low GAB1 expression. Moreover, the association between the co-expression of GRB2/GAB1 and the survival rates were tested by the method of Kaplan-Meier. The Chi-square value by Log Rank (Mantel-Cox) indicated a significant difference among different groups with regard to the conjoined expression status of GRB2/GAB1 (Figure 6E and F). The results by pairwise comparisons showed that the statistically significant difference of disease-free survival and overall survival existed between GRB2-high/GAB1-high patients and any of other three groups (both P<0.001). In all four groups, GRB2-high/GAB1-high patients had the poorest prognosis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Disease-free survival and overall survival curves for two groups defined by low and high expression of GRB2 (A and B) and GAB1 (C and D), and for four groups defined by combined expression of GRB2 and GAB1 (E and F), in patients with HCC.

The patients with high GRB2 and GAB1 expression had a significantly shorter 5-year overall and disease-free survival rate than those with low GRB2 and GAB1 expression (both P=0.008). In addition, the results by pairwise comparisons showed that the statistically significant difference of overall and disease-free survival existed between GRB2-high/GAB1-high patients and any of other three groups (both P<0.001). In all four groups, GRB2-high/GAB1-high patients had the poorest prognosis.

https://doi.org/10.1371/journal.pone.0085170.g006

Furthermore, in a multivariate Cox model, including serum AFP, tumor stage, tumor grading, presence of cirrhosis, gender, age, GRB2 expression, GAB1 expression and combined GRB2/GAB1 expression, we found that GRB2 expression (both P=0.01, Table 2), GAB1 expression (both P=0.01, Table 2) and combined GRB2/GAB1 expression (both P=0.001, Table 2) were independent poor prognostic factors for both 5-year disease-free survival and 5-year overall survival in HCC.