Abstract
Background: During tumorigenesis of gastrointestinal stromal tumors (GISTs), the most frequent changes are reported to be gain-of-function mutations in the C-KIT proto-oncogene. However, we speculated that additional genetic alterations are required for the progression of GISTs. Patients and Methods: Using 15 cases diagnosed with GISTs, we searched for novel indicator genes by microarray analyses using an Oligo GEArray® PI3K-AKT Signaling Pathway Microarray Kit. In addition, we analyzed the mutational status of C-KIT and the proliferation status indicated by the Ki-67 index. Results: The tumor localizations of the 15 GISTs were as follows: 8 in the stomach; 2 in the small intestine; 2 in the mesentery; 1 in the duodenum; 1 in the rectum; and 1 in liver. Regarding the C-KIT gene analysis, mutations in exon 11 were detected in 11 out of 13 patients. In 1 out of the 13 patients, mutations were detected in both exons 11 and 13. No genetic abnormalities were identified in 1 patient. The Ki-67 labeling indices were significantly lower for the low-risk and intermediate-risk groups than for the high-risk group (p=0.0440). No specific genes were overexpressed in the >1% Ki-67 group. Regarding the primary lesion sites, the following 6 genes were overexpressed in tumors in the stomach: RBL2, RHOA, SHC1, HSP90AB1, ACTB and BAS2C. Conclusion: Gene analysis is currently only useful for diagnostic assessment and predicting therapeutic effects. However, it may be possible for new malignancy-related factors to be identified by comparing and investigating gene expression levels and other factors using such analyses.
Gastrointestinal stromal tumors (GISTs) are mesenchymal tumors that arise in the gastrointestinal tract, and account for about 1% of gastrointestinal tumors (1). During tumorigenesis of GISTs, the most frequent changes are reported to be gain-of-function mutations in the C-KIT proto-oncogene. These mutations can result in autophosphorylation, namely KIT ligand-independent kinase activity (2). Approximately 95% of GISTs express the receptor tyrosine kinase KIT, which is useful for distinguishing these tumors from other sarcomas that develop in the abdomen (3). In recent years, it has been established that 75-80% of GISTs harbor mutations in the KIT gene, and that the resulting KIT mutants play important roles in the development of these tumors (3). Furthermore, therapeutic targeting of KIT mutants with the tyrosine kinase inhibitor imatinib (Gleevec, Glivec; Novartis Pharma AG, Basel, Switzerland) has proven to be an effective treatment for patients with advanced unresectable GISTs. However, the efficacy varies depending on the locations of the mutations (4). Recently, approximately 5% of GISTs were found to be characterized by mutations in the related receptor tyrosine kinase platelet-derived growth factor receptor alpha (PDGFRA) exons 12 and 18 (5, 6). Very recently, Lasota et al. (7) reported PDGFRA exon 14 mutations in 11 of 200 GISTs that were negative for mutations in C-KIT exons 9, 11, 13 and 17 and PDGFRA exons 12 and 18. However, approximately 5% of all GISTs do not have detectable mutations in either the C-KIT or PDGFRA genes.
Although KIT gene mutations specific to GISTs have been used as important diagnostic indicators, histological factors such as tumor diameter and tumor cell mitosis/density have mostly been used to assess GIST malignancy (8). Therefore, we speculated that additional genetic alterations are required for the progression of GISTs. However, malignancy has not been assessed based on genetic abnormalities. In the present study, we aimed to identify novel indicator genes by microarray analyses using an Oligo GEArray® PI3K-AKT Signaling Pathway Microarray Kit. In addition, we analyzed the mutational status of C-KIT and the proliferation status indicated by the Ki-67 index.
Fletcher's GIST malignancy classification system.
Patients and Methods
Patients and tumor tissues. The study comprised 15 patients (7 men and 8 women) diagnosed with GISTs between July 2005 and July 2008, and included 2 patients with recurrent GISTs. The specimens excised by surgical resection were histopathologically assessed and classified into four groups using Fletcher's GIST malignancy classification system (Table I) (8).
The risk grades of the GISTs were evaluated according to Fletcher et al. (8). Tumors of <2 cm in diameter with 0-4 mitoses/50 high power fields (HPF) were considered to be very low risk, while tumors of 2-5 cm in diameter with 0-4 mitoses/50 HPF were considered to be low risk. Tumors of <5 cm in diameter with 6-10 mitoses/50 HPF or tumors of 5-10 cm in diameter with 0-4 mitoses/50 HPF were considered to be intermediate risk. Tumors of >5 cm in diameter with >5 mitoses/50 HPF, tumors of >10 cm in diameter with any number of mitoses or tumors with >10 mitoses/50 HPF were considered to be high risk. Tumors with necrosis were also classified as high risk.
C-KIT gene analysis. Genomic DNA was extracted from formalin-fixed paraffin-embedded tumor tissues using standard proteinase K digestion and phenol/chloroform extraction methods, and used for subsequent molecular analyses. To avoid contamination from normal tissues, only tumor tissues were scraped and obtained from each paraffin section, by reference to serial sections stained with hematoxylin and eosin.
Exons 9, 11, 13 and 17 of C-KIT were amplified by PCR using previously published primer sets (Table II). Each of the amplified fragments was purified from a polyacrylamide gel after electrophoresis, and subjected to direct sequencing. All the sequencing reactions were carried out in both the forward and reverse directions.
Immunohistochemistry of Ki-67. Resected specimens were fixed with 10% (v/v) formaldehyde and embedded in paraffin blocks. Immunohistochemical staining was performed on sections using a standard avidin-biotin-peroxidase complex (ABC) method. Briefly, 4-μm sections were cut from the paraffin blocks, deparaffinized and treated with a 0.3% (v/v) H2O2/methanol solution for 30 min at room temperature to block endogenous peroxidase activity. After rehydration through a graded ethanol series, the sections were microwaved in 10 mM phosphate citrate buffer (pH 8.0) at 90°C for 15 min, and cooled to 30°C. After rinsing in 0.1 M phosphate-buffered saline (PBS; pH 7.4), nonspecific binding sites were blocked by incubation with 10% (v/v) normal horse serum for 30 min. Next, the sections were incubated with an anti-human Ki-67 murine monoclonal antibody (Dako, Carpinteria, CA, USA) as the primary antibody overnight at 4°C. Subsequent immunohistochemical staining steps were performed using an ABC system (Vectastain; Vector Laboratories Inc, Burlingame, CA, USA). Briefly, incubation with biotinylated anti-rabbit IgG as the secondary antibody was carried out for 30 min. The chromogen utilized was 3,3′-diaminobenzidine tetrahydrochloride, which was applied as a 0.02% (w/v) solution containing 0.005% (v/v) H2O2 in 50 mM ammonium acetate/citric acid buffer (pH 6.0). The sections were lightly counterstained with hematoxylin.
TUNEL staining. TUNEL staining was carried out using Apoptag® Plus Peroxidase In Situ Kit (Chemicon, Temecula, CA). Briefly, deparaffinized hydrated esophageal biopsy sections were treated with proteinase K (20 μg/ml) in 0.05 M PBS (pH 7.4) at room temperature for 15 min. After washing with distilled water, the sections were incubated with 3% H2O2 for 5 min. Each slide was then covered with equilibration buffer at room temperature for 1 h, incubated with working-strength terminal transferase (TdT) enzyme at 37°C for 1 h in a humidified chamber, and washed with 0.05 M PBS. The sections were then incubated with a peroxidase-conjugated anti-digoxigenin antibody at room temperature for 30 min in a humidified chamber, washed with 0.05 M PBS and incubated with diaminobenzidine (DAB) solution for 6 min. After washing with distilled water, the sections were counterstained with Mayer's hematoxylin (Sigma Chemical Co, St. Louis, MO, USA) and coverslipped.
Oligo GEArray. RNA was extracted from each frozen specimen using an acid guanidinium-phenol-chloroform method. Microarray analyses were conducted using an Oligo GEArray® PI3K-AKT Signaling Pathway Microarray Kit (SuperArray, Frederick, MD, USA). For each total RNA sample (2 μg), a TrueLabeling-AMPTRM 2.0 Linear RNA Kit (SuperArray) was used to synthesize cDNA, and an IVT enzyme mix was used to prepare biotin-labeled cRNA. The samples were hybridized with membranes containing spots for biotin-labeled cRNA and 122 DNAs (Table III) overnight at 60°C. The hybridized membranes were placed in a CDP-Star substrate solution for 5 min at room temperature, and luminescent signals were detected using a charge-coupled device camera system. The GEArray® Expression Analysis Suite (SuperArray) was used to analyze the results. Since the tumor diameter was smallest in Case 1, the expression levels were compared with those in Case 1. The cases were divided into two groups with respect to their median gene expressions to compare the risk classifications, primary tumor sites and Ki-67 labeling indices.
Statistical analysis. The Ki-67 labeling indices were calculated as the percentages of cells with nuclear staining of Ki-67 in 10 HPFs (HPF; ×400). The TUNEL indices were calculated as the numbers of TUNEL-positive cells among more than 1.000 cells counted. The relationships between the Ki-67 labeling indices and other parameters were determined by the analysis of variance method.
Results
Clinicopathological characteristics. The clinicopathologic characteristics of the patients are shown in Table IV. The mean age of the 15 patients was 59.3 years (range, 25-80 years). The GIST localizations were as follows: 8 in the stomach; 2 in the small intestine; 2 in the mesentery; 1 in the duodenum; 1 in the rectum; and 1 in liver. The mean tumor size (maximum diameter) was 6.3 cm (range, 0.7-16 cm). According to the risk categories, 6 cases were classified as high risk, 4 as intermediate risk, 4 as low risk and 1 as very low risk.
Sequences of the primers used for mutational analyses of the C-KIT gene.
C-KIT gene analysis. Among the 15 cases, mutations in C-KIT exons 9, 11, 13 and 17 were analyzed in 13 patients, excluding 2 patients who underwent emergency surgery (Table IV). In 11 out of the 13 patients, mutations in exon 11 were detected. In 1 out of the 13 patients, mutations were detected in both exons 11 and 13. One patient had no genetic abnormalities. No correlations existed between the C-KIT genetic mutations and the risk classifications.
Immunohistochemistry of Ki-67. The Ki-67 labeling indices were 0.89 in the low-risk group included very-low-risk group, 3.54 in the intermediate-risk group and 5.31 in the high-risk group. The Ki-67 labeling index was significantly lower for the low-risk and intermediate-risk groups than for the high-risk group (p=0.0440).
TUNEL staining. In all patients for whom immunostaining was conducted, the apoptosis index evaluated by TUNEL staining was <1% and no correlations with the risk classifications were identified.
Oligo GEArray. Among the 15 patients, the gene expressions were analyzed in 11 patients, excluding Cases 1, 2, 14 and 15. A total of 122 genes in the 11 specimens were expressed at low or high level. The cases were divided into two groups with respect to their Ki-67 labeling indices (cutoff value, 1%). The following 30 genes were also expressed as higher in the >1% Ki-67 group: RPS27A, ADAR, APC, AXIN1, CCND1, CDC42, CDKN1B, CUTL1, EIF4B, FOS, FZD1, HSPB, HSP90AA2, ILK, INPPL1, ITGB1, MAP2K1, MYD88, NFKB1, NFKBIA, PIK3R1, RBL2, RHOA, RPS6KA1, TSC2, YWHAH, ZFYVE21, B2M, HSP90AB1, and ACTB. Among these 30 genes, PIK3RI and B2M were underexpressed in the ≤1% Ki-67 group (Figure 1). These gene expressions were analyzed in relation to the risk classifications, and the following 5 genes were overexpressed in the three high-risk patients: EIF4A1, HSPB, ITGB, RHOA, and WASL. Among these 5 genes, only WASL was underexpressed in the low-risk patients (Figure 2). Regarding the primary lesion sites, the following 6 genes were overexpressed in tumors in the stomach: RBL2, RHOA, SHC1, HSP90AB1, ACTB, and BAS2C (Figure 3).
Discussion
GIST is a generic term for mesenchymal tumors arising in the gastrointestinal tract. Rosai et al. (9) proposed a classification system and diagnostic and therapeutic criteria are currently being established, but there remains room for discussion.
In the present study, mutations in exon 11 were highly detected and the Ki-67 labeling index was significantly higher for the high-risk group. Regarding the results using Oligo GEArray, positive association between gene expression and the Ki-67 labeling index were found in PIK3RI and B2M. The association between gene expressions and the risk classifications was found only in WASL. Regarding the primary lesion sites, the six genes were overexpressed in tumors of the stomach.
In GIST patients with exon 11 KIT mutations, previous studies have reported high levels of imatinib sensitivity (10, 11). Most patients in the present study showed exon 11 KIT mutations, and imatinib was expected to be effective (12). However, although KIT gene analyses may be part of the standard diagnostic procedure for GISTs, they may not indicate imatinib sensitivity. In fact, C-KIT gene analyses were not found to be a better indicator of malignancy than the risk classifications based on tumor diameter and mitosis (8). In terms of immunostaining, TUNEL staining was not useful for assessing malignancy, while the Ki-67 labeling index was higher for the high-risk group, similar to the findings of previous studies (13, 14). Our findings show that the Ki-67 labeling index is an indicator of malignancy in GIST.
The present analyses identified two candidate genes, PIK3R1 and B2M, that were overexpressed in the >1% Ki-67 group and underexpressed in the ≤1% Ki-67 group. PIK3R1 is a gene involved in cellular proliferation, differentiation, apoptosis and cytoskeleton reconstruction (15). The risk classification system devised by Fletcher et al. (8) takes into account the tumor diameter and mitosis. We found that WASL was overexpressed in the high-risk group and underexpressed in the low-risk group. WASL is a gene involved in cellular structure and movement, and may be useful in risk-assessment criteria (16). In terms of the primary lesion sites, SHC1 was overexpressed in stomach lesions and underexpressed in other lesions. SHC1 is a gene that facilitates MAPK activation and cellular proliferation regulation (17). In the present study, the kinase activities varied between the stomach and other regions of the gastrointestinal tract, and differences in malignancy and therapy may exist with respect to the location of the primary lesion.
Target RNAs analyzed in this study.
Ki-67 labeling indices and gene expressions using the Oligo GEArray. The cases were divided into two groups with respect to their Ki-67 labeling indices (cutoff value, 1%). The expressions of the following 30 genes were higher in the >1% Ki-67 group: RPS27A, ADAR, APC, AXIN1, CCND1, CDC42, CDKN1B, CUTL1, EIF4B, FOS, FZD1, HSPB, HSP90AA2, ILK, INPPL1, ITGB1, MAP2K1, MYD88, NFKB1, NFKBIA, PIK3R1, RBL2, RHOA, RPS6KA1, TSC2, YWHAH, ZFYVE21, B2M, HSP90AB1, and ACTB. Among these 30 genes, PIK3RI and B2M were underexpressed in the ≤1% Ki-67 group.
Risk classifications and gene expression levels using the Oligo GEArray. The following 5 genes were overexpressed in the three high-risk patients: EIF4A1, HSPB, ITGB, RHOA, and WASL. Among these 5 genes, only WASL was underexpressed in the low-risk patients.
Origin and gene expressions using the Oligo GEArray. Regarding the primary lesion sites, the following 6 genes were overexpressed in tumors in the stomach: RBL2, RHOA, SHC1, HSP90AB1, ACTB, and BAS2C .
Patients' characteristics, risks, gene alterations, TUNEL indices and Ki-67 indices.
By comparing and analyzing various factors and microarray gene analyses, new indicators for malignancy may be identified in the future. Since the present analyses were based on mRNA, further studies are required to analyze the protein expressions.
In conclusion, gene analysis is currently only useful for diagnostic assessment and predicting therapeutic effects. However, new malignancy-related factors may be identified by comparing and investigating gene expression levels and other factors using such analyses.
Acknowledgements
This work was supported in part by a Grant-in-Aid for Scientific Research (No. 21591691).
- Received February 10, 2010.
- Revision received May 21, 2010.
- Accepted May 26, 2010.
- Copyright© 2010 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved