Dysregulated miR expression has been reported across a broad spectrum of human cancers, suggesting their involvement in carcinogenesis in general. Several studies have demonstrated that miRs can exhibit oncogenic as well as tumor-suppressive activity. A better understanding of miR expression alterations in human cancer may lead to improved diagnosis and management, ultimately leading to improved survival and quality of life for cancer patients.
Colorectal cancer represents a well-defined complication of longstanding colonic inflammation in patients with IBD. The risk of cancer increases with the extent and duration of colonic inflammation. Moreover, dysplasia identified in inflamed colonic mucosae correlates strongly with multiple concomitant foci of dysplasia or frank carcinoma elsewhere in the same colon. Therefore, dysplasia in IBD often leads to total colonic resection, performed prophylactically to prevent the morbidity and mortality associated with colon cancer per se
. Unfortunately, histopathologic diagnosis of dysplasia is problematic. Morphologic alterations associated with chronic inflammation make dysplasia difficult to diagnose and prone to subjective interpretation3
. Current guidelines recommend that any lesion suspicious for dysplasia must be evaluated by two independent pathologists who specialize in gastrointestinal pathology. It is therefore crucial to develop a better understanding of malignant progression in IBD, in the hope that this improved understanding will translate into a positive impact on the early diagnosis, management and individualization of medical and surgical care for IBD patients.
To date, there have been no published reports describing the involvement of miRs in the pathogenesis, diagnosis or prognosis of colonic neoplasia arising in the setting of IBD. In the current study, we set forth to determine global miR expression alterations during neoplastic progression in IBD. In order to detect the earliest changes associated with neoplastic transformation, we first compared global miR expression levels in IBD-associated dysplasia vs. chronically inflamed IBD specimens. This analysis identified aberrant expression of several miRs (). These miRs are associated with neoplastic progression in IBD, and some may even play an etiologic role. Therefore, given the current diagnostic difficulties with correctly identifying dysplasia in IBD, it is attractive to speculate that at least some of these miRs may be applied to the early diagnosis or prediction of neoplastic progression in IBD.
We selected miR-31 for further studies because it was one of the most differentially expressed miR species between IBDs and IBDNs. Additionally, previous work from other laboratories found miR-31 to be upregulated in sporadic colorectal carcinomas 18–20
. However, its expression in IBD and IBD-associated cancers or dysplasias has not been previously reported.
To date, there are no studies describing the relationship between miR expression and colonic anatomic location. With miRs emerging as a new class of potential biomarkers for cancer diagnosis, this relationship becomes especially important in the context of differences between miR expression in right vs. left-sided colon, since current screening methods often miss right-sided lesions. Therefore, we sought to determine whether miR-31 expression levels were affected by colonic anatomic location. By systematically assessing the expression of miR-31 in 55 colonic mucosae specimens, we established that miR-31 expression levels do not differ among different anatomic locations. This finding suggests that, at least in the case of miR-31, dysregulated expression is indicative of a pathologic condition, rather than a location bias. We subsequently sought to comprehensively determine miR-31 expression levels in a large independent cohort of specimens collected at different stages of IBD-related neoplastic transformation. Our aim was to dissect changes in miR-31 expression as they occurred along the inflammation-cancer axis. We found that the expression of miR-31 becomes progressively higher during progression from normal colonic mucosae, to inflamed tissue, with the highest level in cancer. The increase in miR-31 expression was independent of age, sex, anatomic location, stage, or differentiation grade (data not shown). This independence of clinical parameters suggests that miR-31 may be associated with inherently unique molecular pathways underlying IBD-associated colon carcinogenesis. Of note, a statistically significant increase in miR-31 expression was observed at both the dysplastic and cancerous stages vs. chronically inflamed tissue in patients with IBD. This relatively early upregulation of miR-31 suggests that early changes associated with neoplastic transformation can be accurately diagnosed by elevations in miR-31. While it is difficult at this point to directly determine whether miR-31 plays an etiologic role in IBD-associated malignant transformation, its expression levels can potentially already be used to diagnose or predict early carcinogenesis.
Previous studies reported that miR-31 is upregulated in sporadic colorectal cancers. However, there have been no reports regarding miR-31 or any other miR expression in IBD-associated colon cancers. Our data demonstrates that miR-31 expression levels are significantly higher in IBD-associated cancers than in their sporadic colorectal counterparts. This result confirms that while there are similarities between IBDNs and sporadic colorectal cancers, there are also several important differences, including miR-31 expression levels. Increased miR-31 expression in IBDNs vs. sporadic colorectal cancers may also be due to the marked inflammation associated with IBDNs. This finding may have great potential for clinical application, since miR-31 expression may ultimately be applied to the critical decision of colonic resection (in the case of IBD-related neoplasia) vs. localized polypectomy/mucosal resection (in the case of sporadic neoplasia). Future studies in independent cohorts are required to evaluate and validate the performance of miR-31 qRT-PCR levels in this context.
The ability to isolate nucleic acids suitable for molecular analyses from serum samples provides a powerful venue for biomarker discovery and application. Recent studies have demonstrated that due to their small size, as well as to their containment within protein complexes, miRs are more resistant to degradation than are mRNA or DNA, and therefore potentially more suitable as blood-based biomarkers for early cancer detection 21
. Interestingly, results of TaqMan low-density array qRT-PCR profiling did not detect miR-31 in human plasma from a healthy donor21
. Since its normal level is virtually zero, it is tempting to speculate that if miR-31 was released into the blood by cellular breakdown or apoptosis in IBD patients, it would be stable and easily detected.
The progressive increase we observed in expression levels from normal to chronically inflamed mucosae to dysplasia and frank cancer mirrors the IBD-related adenocarcinoma sequence. Determining the actual functions of miR-31 in colon inflammation and carcinogenesis may provide valuable new insights crucial to our understanding of biological events that promote evolution from chronic inflammation to carcinoma in general.
The requirement for neovascularization is universal to all solid tumors, including colon cancers. Pathological angiogenesis has been shown to play a pivotal role in tumor progression and metastasis22
. Experimental and clinical studies have demonstrated that angiogenesis also plays a pivotal role in inflammatory bowel disease 23
. Studies have shown that inhibition of angiogenic response in inflammatory bowel disease attenuates disease pathology24, 25
. The mechanism of tumor angiogenesis is primarily mediated by hypoxia through chronic activation of the hypoxia inducible factor (HIF) pathway, leading to the production of vascular endothelial growth factor and basic fibroblast growth factor26–28
. Interestingly, in our study, miR-31 negatively regulates FIH1. FIH1 is a hydroxylase that catalyzes the post-translational modification of HIF. Hydroxylation of HIF1 both causes its degradation and limits its activity. Sustained increased expression of miR-31 during chronic inflammation could therefore translate into decreased FIH1 repressor activity, allowing for increased transcription of HIF1-regulated genes. While this manuscript was in preparation, a new functional study describing the FIH1 regulation in head and neck squamos cell carcinoma by mir-31 was published. The authors also demonstrated that ectopic miR-31 expression increased oncogenic phenotypes of head and neck squamos cell carcinoma cells under normoxic conditions in cell culture or tumor xenografts.
The current study brings several critical advances: 1) this is the first study to assess the expression of any miRs in colonic dysplasia and cancer arising in IBD; 2) this is the first report assessing the expression of any miR in distinct colonic anatomic locations, with the purpose of identifying potentially confounding changes related to location rather than pathology per se; 3) this is the first report to identify miR-31 as a significantly differentially expressed miR species along the axis of inflammation to cancer in IBD; 4) this study demonstrates that miR-31 can accurately discriminate between IBD-associated and sporadic colonic neoplastic lesions; and 5) this study suggests that miR-31 is a potential biomarker for the early detection or prediction of dysplasia and colorectal carcinoma developing in patients with IBD. Notably, all of these findings were demonstrated in a large cohort of patient specimens. Finally, this study identifies FIH1, a repressor of the pro-angiogenic hypoxia-inducible factor HIF1, as a direct target of miR-31 in colon cancer cells.