Final Progress Report
Proposal No. IBD-0065R
Principal Investigator: Randall F. Holcombe, M.D.
Applicant Organization: University of California, Irvine (U.S.A.)
Project Title: Wnt signaling in inflammatory bowel disease-related colon cancer
Period of Award: November 1, 2003 - October 31, 2005
This report summaries activities for BMRP grant IBD-0065R. Some additional work related to genetic analysis of the axin1 locus in patients with IBD and the microarray studies outlined in the initial proposal remains to be completed. These studies will be completed by 7/1/06 utilizing intramural funding available to the PI.
A. Achievements
Specific Aim 1A: Test the hypothesis that the expression of frizzled (Fz) receptors subtypes, frizzled inhibitory proteins, disheveled (DVL) proteins and selected Wnt ligands is altered in IBD-associated colon cancer.
Samples of colon cancer, and in most cases matched normal colonic epithelium, arising from patients with IBD have been obtained through the NIH-sponsored cooperative tissue network based at Vanderbilt University and through the Pathology Department at the University of California, Irvine. These have been analyzed in detail. The expression of Fz receptors and DVL protein family members shows interesting and significant differences in colon cancer arising in patients with IBD compared to sporadic colon cancer. This work has been completed and was submitted for publication consideration to Inflammatory Bowel Diseases. The manuscript, “Expression of Wnt pathway Components Frizzled and Disheveled in Colon Cancer Arising in Patients with Inflammatory Bowel Disease” was not accepted and several revisions were suggested. A copy of the submission has been previously submitted to the Broad Medical Foundation. The manuscript is currently being revised for resubmission.
Manuscript abstract:
Mutations in apc which lead to activation of the Wnt signaling pathway are a hallmark of sporadic colon cancers but occur infrequently in colon cancers arising in patients with inflammatory bowel disease (IBD). There is evidence, however, that other components of the Wnt pathway may be altered in IBD-related colon cancer. In this study, we examined the expression the Wnt pathway components frizzled (Fz), the cell surface receptor, and disheveled (DVL), a family of cytoplasmic signal transduction molecules, in IBD and IBD-related colon cancer. Paraffin sections of normal and malignant colon tissues were obtained from patients with a history of ulcerative colitis and from controls with sporadic colon cancer. Tissue sections were stained with antibodies directed against Fz1/2 receptors and DVL1, DVL2 and DVL3 and antigen expression visualized by immunohistochemistry. Fz1/2 receptors were minimally expressed in normal IBD mucosa, were not expressed in IBD colon cancer, but exhibited strong expression in dysplastic tissues adjacent to the cancers. DVL1 was not expressed in IBD normal mucosa or normal mucosa from non-IBD patients, but was expressed in all cancers. DVL2 and DVL3 were expressed in all normal mucosa samples tested, were expressed in sporadic colon cancer, but were not expressed in colon cancers arising in IBD patients. The characteristics of Fz and DVL expression in IBD tissues reported here provides evidence of the importance of Wnt signaling in IBD and IBD-related colon cancer and, specifically, the significance of non-APC components of this pathway. Fz may serve as a marker for dysplasia in IBD patients and DVL1 is a potential therapeutic target for IBD-related colon cancer.
Specific Aim 1B: Test the hypothesis that genetic mutations in specific Wnt pathway components other than APC occur in IBD-associated colon cancer.
Significant progress has been made on this Aim. Initial studies have focused on identifying mutational changes in the genes for axin1, axin2 and casein kinase Iε (CK1ε). No mutations in CK1ε were identified that were specific for IBD-related colon cancer. Mutations identified previously in breast cancer by Dr. Bryant, a co-investigator on this project, were not identified. Hot spots for mutations in the axin1 gene have been identified as 3 locations in exon 1, 2 locations in exon 2 and 2 locations in exon 10.
We developed primers for analysis of the specific sequences including two axin1 primer pairs for each of the exon 1 and exon 10 hot spots. Two specific mutations in the coding sequences have been identified at base pairs 821 and 930 (Figure 1). We are concentrating on these mutations and testing all available IBD and sporadic colon cancer specimens in order to define the frequency of occurrence and their specificity for IBD. Numerous colon cancer cell lines are also being tested. We continue to evaluate for mutational changes which may be specific for IBD colon cancer at the other hot spot locations as well. We have also developed primer pairs which generate variably-sized products by PCR depending on the presence or absence of the mutations. This has the potential for future utilization for diagnostic testing if determined reasonably specific for IBD colon cancer.
AxnNTRKO84 (821-ACG / Threonine)
AxnNTCL29/Wt ( 930- ATG / Methionine)
Figure 1. Sequence analysis revealing two specific mutational changes in exons of the axin1 gene.
In the course of this work, we have identified two isoforms of the axin2 gene which have not been reported previously. We have characterized these isoforms molecularly. They arise via alternative splicing at the exon6 splice junction sites producing full length axin2 and a shortened product lacking exon6 (Figure 2). This change is in close proximity to the β-catenin binding domain and therefore we postulate that the different isoforms will have different affinity for β-catenin which may have functional significance.
This work on axin2 has been submitted for publication consideration to Gene Expression. A copy of this submission is appended to this report.
Specific Aim 2: Test the hypothesis that the expression of frizzled receptor subtypes, frizzled inhibitory proteins, DVL proteins and selected Wnt ligands is altered in dysplastic and inflamed intestinal mucosa from patients with IBD.
The clinical protocol was approved by the Chao Family Comprehensive Cancer Center scientific review committee (CTPRMC) as well as the University of California, Irvine institutional review board (IRB). Dr. Albers has identified patients who present for routine colonoscopic screening through the Division of Gastroenterology and the Comprehensive Digestive Diseases Center at UCI. Originally, we obtained tissue with snap freezing at the time of colonoscopic biopsy. We found this did not result in optimal RNA salvage and have changed to a method utilizing RNA-later for tissue collection. This results in higher quality RNA preservation. Also, we have just recently opened the protocol for accrual at the Long Beach VA Medical Center with the assistance of Dr. Jamal because the volume of patients consenting to additional biopsies at UCI Medical Center/Chao Cancer Center has been somewhat low. Additional paired sets of biopsies have been obtained.
The original aims outline multiple tests by immunohistochemistry and in situ hybridization to be performed on these tissues. We decided to focus on Wnt-specific microarray analyses as this will provide more information related to more markers while preserving the scarce RNA resource from small biopsy specimens. An example of the microarray data, from tissue obtained from patient #1, is presented in Figure 3. In patient #1, there was decreased expression in the inflamed tissue of LRP6, a co-receptor for Wnt ligands at the cell surface, and Jun, a component of non-canonical signaling pathways. Increased expression was seen for sFRP2, a secreted inhibitor of Wnt signaling and DVL2. The changes in LRP6 appear to be consistent across different patients. Two other interesting changes have been noted: decreased expression of Wnt2 and myc in inflamed mucosa. Based on the preliminary data, it appears that Wnt signaling may be downregulated in inflamed mucosa in most patients with IBD compared to non-inflamed mucosa. If augmentation of Wnt signaling is important in the development of IBD-colon cancer, yet to be directly proved though abnormalities in the expression of various pathway components have been defined, then the processes involved in carcinogenesis in the IBD colon may be quite distinct from those involved in inflammation. Additional microarray analyses are in process. When completed, information regarding any publication submission will be provided to the Foundation.
B. Significance to IBD
These studies represent a unique avenue of exploration into the role of Wnt signaling in the pathogenesis of IBD-related colon cancer. The data derived under this proposal suggest that Wnt signaling may be activated in the “normal” colonic mucosa. This could result in increased cell turnover, increased proliferation and a propensity for malignant transformation. Alternatively, the Wnt activation may be the result on ongoing inflammatory stimuli present in individuals with IBD that may be absent in non-affected individuals. If the former, identifying ways to block Wnt signaling will be important in order to prevent the eventual development of colon cancer. If the latter, components of the Wnt pathway may serve as sensitive markers of disease activity. The identification of specific mutations which arise in IBD-associated colon cancer may lead to clues regarding the pathogenesis of cancer in this population.
C. Lay Summary
Patients with inflammatory bowel disease have an increased likelihood of developing colon cancer. An important signaling pathway which is involved in the development of colon cancer in patients without IBD is the Wnt signaling pathway. This pathway becomes dysregulated and promotes the change from normal tissue to cancerous tissue. Under this research funded by the Broad Medical Research Foundation, we have studied the activity of the Wnt signaling pathway in patients with IBD. Abnormalities in cell surface receptors involved in the Wnt pathway were found in pre-malignant IBD colon tissue which may be useful in identifying patients at highest risk for colon cancer development. Molecular studies reveal that the Wnt signaling pathway may be downregulated, or turned-off, in areas with significant inflammation, suggesting that the processes involved in IBD inflammation may be very distinct from the processes involved in cancer development, or that colon cancer arising in the setting of IBD has molecular origins distinct from colon cancer in patients without this disease.