- Top of page
- Material and Methods
- Disclosure Statement
The invasion process is a crucial step for pancreatic ductal adenocarcinoma (PDAC); however, the genes related to invasion remain unclear. To identify specific genes for the invasion process, we compared microarray data for infiltrating cancer and PanIN-3, which were harvested from an individual PDAC patient by microdissection. Furthermore, immunohistochemical, coimmunoprecipitation and invasion analyses were performed to confirm the biologic significance of molecules identified by expression profile. In the present study, we focused on MUC16 and mesothelin among 87 genes that were significantly upregulated in infiltrating components compared to PanIN-3 in all PDAC patients, because MUC16 was the most differently expressed between two regions, and mesothelin was reported as the receptor for MUC16. Immunohistochemical analysis revealed that MUC16 and mesothelin were expressed simultaneously only in infiltrating components and increased at the invasion front, and binding of MUC16 and mesothelin was found in PDAC by immunoprecipitation assay. The downregulation of MUC16 by shRNA and the blockage of MUC16 binding to mesothelin by antibody inhibited both invasion and migration of pancreatic cancer cell line. MUC16 high/mesothelin high expression was an independent prognostic factor for poor survival in PDAC patients. In conclusion, we identified two specific genes, MUC16 and mesothelin, associated with the invasion process in patients with PDAC. (Cancer Sci 2012; 103: 739–746)
For most patients with pancreatic ductal adenocarcinoma (PDAC), the diagnosis is made at an advanced stage; the survival rate for these patients is dismal because PDAC has a propensity for early local invasion and vascular dissemination. The genetic and biochemical determinants of the process of invasion and metastasis in PDAC are still largely unknown.
Pancreatic ductal adenocarcinoma appears to arise from histologically well-defined precursor lesions in the ducts of the pancreas, called pancreatic intraepithelial neoplasms (PanIN).[3, 4] PanIN are graded based on their degree of architectural and nuclear atypia and are categorized into a four-tier classification, including PanIN-1A, 1B, 2 and 3. PanIN-3 lesions demonstrate widespread loss of nuclear polarity, nuclear atypia and frequent mitoses, and whereas cancerous cells break through the basement membrane, they evolve into infiltrating adenocarcinoma. The invasion process is the crucial step in PDAC because cancer cells that invade the vasculature, or lymphatic or neural vessels, can progress further to metastasis only after obtaining infiltrating status. In the present study, we identified specific molecular markers associated with invasion in PDAC, which might be useful not only as early diagnostic markers but also as new therapeutic targets for patients with PDAC.
Several molecular markers, including tissue plasminogen activator, artemin and RhoGDI2, have been reported to be associated with invasion in PDAC. However, some of these molecular markers are of little clinical value as therapeutic targets for patients with PDAC because these genes are also expressed in normal pancreatic tissues or other normal organs.[6-8] In this study, we first used a gene expression profiling technique to identify the specific genes that are differentially expressed between infiltrating cancer cells and PanIN-3 cells, which were harvested from an individual patient by laser microdissection. Based on our gene expression array data, clinical and biologic implications of MUC16 and mesothelin expression were further explored.
- Top of page
- Material and Methods
- Disclosure Statement
We first identified genes specific to the invasion process in PDAC using microdissection and gene expression profiling techniques. In this study, we compared microarray data of infiltrating cancer and PanIN3, which were harvested from an individual PDAC patient, to exclude the difference in original gene expression among individuals. Then, we were able to identify similar genes that were differently expressed between infiltrating cancer and PanIN-3 in all five patients.
Among the identified upregulated genes, we focused on MUC16 because its expression in the infiltrating cancer was substantially higher than that in the PanIN-3 cells. We also focused on mesothelin in the list, because it was reported to be a ligand receptor of MUC16. Their interaction has been postulated to play an important role during tumorigenesis and metastasis in ovarian cancer.[24, 25] Rump and colleagues reported that the binding of MUC16 and mesothelin expressed by cancer cells mediates heterotypic cell adhesion and might contribute to the metastasis and invasion of ovarian cancer.
In the present study, immunohistochemical analysis revealed that MUC16 and mesothelin were expressed in the infiltrating cancer cells but not in the PanIN-3 cells or normal pancreatic tissues, consistent with the results of gene expression profiling. Furthermore, fluorescence immunohistochemistry showed that MUC16 and mesothelin were expressed simultaneously in the PDAC cells.
MUC16 encodes the CA125 antigen and is a membrane-bound mucin protein with a high molecular weight between 2.5 and 5.0 million daltons. Its proposed structure comprises an N-terminal domain of >22 000 amino acid residues that are presumably heavily glycosylated, a central domain containing up to 60 glycosylated repeat sequences constituting the tandem repeats characteristic of mucins, and a C-terminal domain composed of a transmembrane domain and a short cytoplasmic tail with possible phosphorylation sites. Few reports have described the expression of MUC16 in cancers. In this study, using immunohistochemistry, we detected the expression of MUC16 in 94 of 103 PDAC cases (91%).
The mesothelin gene encodes a 71-kDa precursor protein that is processed into the 40-kDa glycosylphosphatidylinositol-anchored membrane glycoprotein, mesothelin and a 31-kDa fragment called megakaryocyte potentiating factor.[29, 30] Mesothelin expression in normal human tissues is limited to mesothelial cells lining the pleura, pericardium and peritoneum, and the protein is also expressed by a variety of solid tumors, including ovarian cancer, malignant mesothelioma, lung cancer and PDAC.[31, 32] Mesothelin expression reportedly conferred chemoresistance and a poorer clinical outcome in ovarian cancer patients.
We found that the coexpression of MUC16 and mesothelin was also increased at the invasion front (n = 48), compared to that in the main tumor in several PDAC tissues, and, then, MUC16 high/mesothelin high expression in PDAC was significantly associated with large tumors, serosal invasion, invasion of other organs and lymphatic permeation. These results indicate that these molecules seem to be involved in invasion and migration of pancreatic cancer cells. Recent reports show the role of MUC16 in ovarian cancer tumorigenesis,[34, 35] and it has been noted that MUC16 regulates cell growth, invasion and metastasis in epithelial ovarian cancer. However, another report indicates the opposite concept, that downregulation of MUC16 inhibits invasion and migration due to the suppression of epithelial to mesenchymal transition in ovarian cancer cells. Thus, the role of MUC16 in ovarian cancer cell invasion and migration is still controversial and no report regarding the role of MUC16 on pancreatic cancer cell invasion and migration has yet appeared.
To examine the role of interaction of MUC16 and mesothelin on pancreatic cancer invasion and migration, we investigated whether shRNA and blocking antibodies for MUC16 suppress invasion and migration of pancreatic cancer cells. We investigated the expression of MUC16 and mesothelin by RT-PCR, western blotting and immunocytochemistry in eight pancreatic cancer cell lines (PK9, PANC1, MIAPaCa2, AsPC1, BxPC3, Capan-1, Capan-2 and PK1). By RT-PCR, both MUC16 and mesothelin mRNAs were detected in five cell lines, including PK9, AsPC1, BxPC3, Capan-2 and PK1. Using western blotting and immunocytochemistry, the strongest positive expressions of both MUC16 and mesothelin were found in PK9. Therefore, in the present study, we used only PK9 cell line for biological experiments. The blockage of the interaction between MUC16 and mesothelin suppressed invasion and migration of pancreatic cancer cells, suggesting that MUC16 binding to mesothelin is important for cell invasion and migration in pancreatic cancer cells.
Furthermore, we focused on the survival of patients with MUC16 high and mesothelin high expression because coexpression of these two genes is obviously correlated to the invasion of PDAC, and MUC16 high/mesothelin high expression was an independent prognostic factor for poor survival. We examined whether there are any differences in survival between the MUC16 high/mesothelin high group and the MUC16 high/mesothelin low group or MUC16 low/mesothelin high group. However, these groups were very small (n = 11), and larger groups of patients are necessary for further study.
The mechanism of overexpression of MUC16 and mesothelin in PDAC has not yet been clarified yet. It is also unclear whether the coexpression of MUC16 and mesothelin was coincidental or the increased expression of MUC16 was associated with an upregulation of mesothelin expression. These issues should be clarified in further studies. Moreover, other molecules in Table 2 besides MUC16 and mesothelin might potentially contribute to the invasion process. In the future, we analyze the roles of other upregulated genes in infiltrating cancer than in PanIN-3 for PDAC patients.
In conclusion, MUC16 and mesothelin are involved in pancreatic cancer cell invasion and migration, and MUC16 and mesothelin clinically represent new prognostic biomarkers for PDAC and might be new therapeutic targets for patients with PDAC, including immunotherapy using a peptide vaccine or monoclonal antibody therapy.