Background
Ulcerative colitis (UC) and Crohn's disease (CD) are forms of inflammatory bowel disease (IBD) which affect an estimated 1.4 million people in the USA and 2.2 million across Europe. UC exclusively affects the large intestine, whereas CD affects the colon in 60% of cases (known as Crohn's colitis), but can also involve other parts of the GI tract [
1]. These common, chronic and debilitating conditions remain incurable, with their aetiology and pathogenesis not clearly understood. Long-term illness greatly increases the risk of colorectal cancer, which causes approximately 15% of all IBD patient deaths [
2]. The onset of cancer in IBD patients is sudden, rapid, and highly aggressive, with a poor prognosis. The occurrence of dysplasia in IBD is widely accepted to be pre-malignant, but the likelihood of progression to cancer is difficult to predict [
3]. Even the distinction between low grade- and high-grade dysplasia provides little indication of disease outcome [
2] and there is, therefore, an urgent need for biomarkers to predict neoplasia in IBD.
Both UC and CD are characterised by mucosal infiltration of inflammatory cells and frequent epithelial damage. This damage can result in destruction of the mucosa and a breach in the barrier that this tissue provides against the luminal milieu. Complete remission of IBD requires both a reduction in inflammation and repair of the damaged epithelium. Inadequate or incomplete repair can result in the formation of a 'leaky barrier' which can in turn perpetuate a vicious cycle of chronic inflammation [
4]. The process of 'healing' areas of the mucosa devastated by inflammation is widely accepted to be a two-stage process comprising 'restitution' and 'regeneration' [
5]. Restitution is characterised by flattening and spreading of the epithelium at the margins of the ulcer, with these cells migrating across the denuded sub-mucosa to cover the damaged area. Following restitution, the regeneration programme requires widespread epithelial cell proliferation and formation of characteristic glandular structures of the intestinal crypts. Although several cytokines and growth factors have been implicated in the restoration of epithelium following injury, the cellular processes underpinning this mechanism remain poorly understood [
5].
Current therapy for IBD, particularly UC, centres on the long-term administration of the non-steroidal anti-inflammatory drug (NSAID) 5-amino salicylate (5-ASA). Shown to be effective in controlling intestinal inflammation in the majority of patients, 5-ASA also reduces colorectal cancer risk in patients with IBD [
6,
7].
Other workers in the field have proposed sodium butyrate, a fermentation product of dietary fibre, as a potential therapy for IBD. Trials using butyrate irrigation led to symptomatic amelioration in UC patients [
8,
9]. Luminal levels of butyrate may be modulated through the dietary intake of fibre and are found in the millimolar range [
10].
Among the most pressing current issues in the clinical management of IBD are a need to further our understanding of intestinal wound healing and to understand the effects of therapeutic modalities on tissue repair as this is crucial for disease remission.
Fascin (fascin-1) is a 55kDa actin-bundling protein which localises to the core actin bundles of spikes and filopodia at the leading edge of migratory cells. It is known to be overexpressed in various cancers and has been shown to increase motility in cells from several different tissues [
11]. Work in this laboratory has shown fascin to be completely absent from the normal colorectal epithelium but widespread in colorectal tumours [
11,
12]. Fascin expression has been shown to correlate with an increased risk of malignant progression and with a poorer disease prognosis making it a potential biomarker in colorectal neoplasia [
12,
13].
As yet, there are no published reports of fascin expression in IBD, but it is our hypothesis that fascin will be involved in tissue repair in IBD. The changes in motile behaviour demonstrated by tumour cells mimic those occurring during tissue repair and require dynamic rearrangements in the actin cytoskeleton, governed by actin-binding proteins such as fascin.
The first aim of this study was to determine the expression of fascin in clinical samples of IBD with particular attention to areas of restitution and regeneration. Samples of resection margins from patients undergoing surgery for diverticulitis were included as these show low-grade mucosal inflammatory infiltration.
Secondly, in order to elucidate the potential modulation of fascin by therapeutic intervention and the subsequent consequences for epithelial repair, we studied the effects of 5-ASA and sodium butyrate on fascin expression and cell motility using an in vitro cell line model.
We now show, for the first time, that fascin is overexpressed in IBD and expression is associated with regions of active mucosal repair. Furthermore, therapeutic modalities for IBD can affect fascin expression and colonic epithelial cell motility.
Discussion
We show here, for the first time, that the actin bundling protein fascin is overexpressed in IBD, and may play an important role in tissue repair following inflammatory damage. Our in vitro findings suggest that therapeutic approaches for IBD could influence fascin expression, subsequently producing unwanted side-effects on tissue repair and thereby influencing achievement of remission.
We have shown previously that fascin is absent from the normal colorectal epithelium but overexpressed in both benign and malignant tumours where it is associated with enhanced cell motility [
11,
12]. Our previous findings in colorectal adenomas, where the tumour stalk frequently showed the strongest fascin immunoreactivity, suggest fascin regulation occurs through epithelial-mesenchymal crosstalk, potentially through inflammatory mediators [
12]. This notion of inflammatory induction of fascin expression in the epithelium is supported here by the widespread expression of fascin in tissue resected from patients with IBD. A positive correlation was observed between the disease activity recorded by the pathologist and the intensity of fascin immunoreactivity further supporting the idea of an inflammation-mediated regulation of fascin expression. Also in keeping with these results is the observation of epithelial fascin expression at the crypt base in tissue from patients undergoing surgery for diverticulitis, which display low-grade inflammation.
The increased proportion and intensity of fascin staining observed in samples of Crohn's compared with UC raises questions about the fundamental difference between these classifications of IBD. As it was only the epithelial immunoreactivity that was scored, this difference cannot be attributed to the ability of Crohn's colitis to affect deeper layers of the bowel wall than UC, but rather resides in the mechanism of fascin regulation.
IBD carries an increased risk of colorectal cancer, resulting in the deaths of 15% of IBD patients [
2]. Malignant progression is highly unpredictable and therefore the identification of prognostic biomarkers of tumourigenic potential is of key importance [
2]. In this study we showed that both the proportion and intensity of epithelial fascin stain significantly correlated with the presence of dysplasia or cancer. Further study, of a larger sample group with long-term clinical follow-up, will be necessary to confirm the validity of fascin as a biomarker for tumorigenesis in IBD. The role of fascin in the malignant progression of sporadic colorectal tumours [
12] does suggest that fascin merits further investigation in the neoplastic transformation of IBD.
One of the most intriguing observations in our study of fascin expression in IBD was the focal expression of fascin in regions demonstrating active tissue repair. Fascin expression was found in all areas of restitution observed, and also in regenerative polyps and branching crypts, suggesting a role throughout the various stages of the repair process. Our previous work has shown that fascin can promote motility in benign, as well as malignant colorectal epithelial cells [
12], and it would have been reasonable to hypothesise that fascin would be expressed in the motile cells undertaking restitution. Furthermore, fascin has been shown to modulate cell adhesion and could therefore also function in the dynamic adhesive changes required for both restitution and crypt fission [
11]. These data suggest that regulation of fascin could be important in achieving and maintaining remission in patients with IBD.
One of the most broadly used treatments for the maintenance of remission in IBD is 5-ASA and previous work has shown this drug to induce apoptosis in HT29 cells in a dose-dependent manner [
18]. We show here for the first time that treatment of the colorectal epithelial cell line HT29 with this drug led to a decrease in fascin expression and a retardation of cellular motility.
Long term use of 5-ASA in patients with IBD is known to reduce cancer risk [
19,
20]. We postulate based on our findings here that this chemopreventive effect may, in part, be explained by the down regulation of fascin expression-with fascin being known to be involved in malignant progression of the colorectal epithelium [
12]. This repression of cellular motility represents a key anti-cancer effect of 5-ASA treatment.
Conversely, considering the potential role of fascin in tissue repair, this would suggest that 5-ASA could hinder wound repair and thus impede remission in patients with active disease. The published evidence is somewhat conflicting as there are reports to suggest that NSAIDs can both impair wound healing [
21] and that 5-ASA can promote intestinal repair [
22]. However, the possibility exists that 5-ASA could have some detrimental effects in IBD patients in addition to its obvious benefits. This paradox in the effect of 5-ASA, and also the role of fascin, could have important consequences for the clinical management of IBD. To this end, our findings suggest that increased luminal levels of butyrate could complement 5-ASA in promoting remission whilst maintaining potent chemopreventive effects against tumorigenesis.
Previous work has suggested that sodium butyrate could have beneficial effects for IBD sufferers [
8,
9]. This fermentation product of dietary fibre is known to be present in the colonic lumen in the millimolar range and levels can be modulated by intake of suitable dietary substrate [
10]. We, and others, have previously shown that butyrate has potent chemopreventive effects against colorectal tumorigenesis through the modulation of differentiation and apoptosis in colorectal epithelial cells [
15,
23,
24].
We show here that butyrate upregulates fascin expression and significantly stimulates colorectal epithelial cell motility. These data suggest that butyrate will aid colonic tissue repair and therefore speed remission in IBD. This short chain fatty acid has already been proposed to promote wound healing in the small intestine [
25]. Although our data suggest that fascin may be involved in tumorigenesis, butyrate has been shown to elicit potent anti-tumour effects. We propose that the combinatorial use of butyrate, or dietary modulation of its levels, could favourably complement 5-ASA use in the clinical management of IBD.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
DQ: Conceived, designed, and coordinated the study and acquired the necessary funding; supervised the laboratory projects of KS & DL; carried out additional immunohistochemistry and all subsequent analyses; carried out some of the in vitro experiments; drafted the manuscript. KS & DL: Carried out the immunohistochemistry and some of the in vitro studies. MP: Contributed to the design and coordination of the study and aided with manuscript preparation. All authors read and approved the final manuscript.