The role of matrix metalloproteinases in chronic inflammatory bowel diseases: new insights and applications.

Inflammatory bowel diseases (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), are chronic, disabling diseases of the gastrointestinal tract. The incidence and prevalence of IBD are increasing world-wide, including developing countries. IBD affects mostly young adults which leads to substantial morbidity and decreased quality of life. Despite extensive research, the etiopathogenesis of IBD is still not completely understood. It is thought that an abnormal mucosal immune response is elicited towards the luminal microbiota in genetically predisposed persons. Accumulating data indicate that matrix metalloproteinases (MMPs) are predominant proteases involved in the onset and persistence of IBD. MMPs are secreted as latent pro-enzymes and require proteolytic cleavage for activation. They are regulated by specific tissue inhibitors of metalloproteinases (TIMPs) and are involved in inflammation and remodeling of the extracellular matrix (ECM). Chronic intestinal inflammation and aberrant tissue remodeling, as seen in IBD, are mediated by an imbalance between MMPs and TIMPs. As a result, many IBD patients suffer from serious complications related to progressive tissue destruction (e.g. ulcers and fistulas) or excessive deposition of collagens resulting in fibrosis.

The disease course of IBD is heterogeneous, patient-specific and difficult to predict due to intermittent phases of active disease and spontaneous or drug-induced quiescent disease. Clinicians rely on a combination of clinical, endoscopic and histological evaluations to follow-up the disease course. However, frequent endoscopic procedures are costly and uncomfortable for the patient. Therefore, there is a need for non-invasive biomarkers to aid clinicians in differential diagnosis and assessment of disease activity or response to treatment. Several blood and fecal markers have been evaluated of which C-reactive protein (CRP) and fecal calprotectin are currently implemented in clinical practice. However, not one single biomarker has been shown to have sufficient sensitivity and specificity to completely override the need for invasive endoscopic evaluations. Several researchers have allocated MMP-9 as a biomarker for disease activity and response to treatment. Serum MMP-9 in combination with fecal calprotectin was associated with imaging/endoscopy-defined inflammation in UC and CD patients. Moreover, fecal MMP-9 was found to correlate with fecal calprotectin and disease activity indices of UC patients.

The introduction of anti-TNF therapy has had a major impact on the therapeutic treatment of IBD patients. Treatment goals evolved from symptomatic remission to mucosal healing of the bowel mucosa and lower hospitalization and surgery rates. However, since one third of the patients do not respond to treatment or they lose response over time, new pharmacological targets are needed. Several studies point towards a causal role of MMP-9 in IBD and recently the therapeutic potential of targeting MMP-9 has been demonstrated in preclinical studies of colitis. As a result, MMP-9 antagonists are currently under investigation in phase 2 clinical trials in UC and CD patients.

The first aim of this PhD project was to further investigate the role of MMPs and TIMPs in IBD. Therefore, we investigated the intestinal mucosal gene expression of all MMPs (n=23) and TIMPs (n=4) in ileal and colonic biopsies isolated from UC and CD patients with active disease before and after first treatment with infliximab (Chapter 3). The expression of many MMPs and TIMPs was increased in IBD patients with active disease, as well as MMP/TIMP ratios. After treatment with infliximab, the expression of MMPs, TIMPs and MMP/TIMP ratios decreased in patients who achieved mucosal healing, whereas MMP and TIMP gene expression levels remained elevated in patients who did not achieve mucosal healing. At the protein level, we found that MMP-9 levels were increased in mucosal tissue from IBD patients with active disease and decreased after treatment with infliximab in patients who achieved mucosal healing.

The second aim was to investigate whether MMP-9 can be used as a biomarker to assess disease activity or response to treatment in IBD patients. Therefore, we investigated MMP-9 levels in serum of UC patients before and after treatment with infliximab (Chapter 4). With zymography analysis, we found that the complex of MMP-9 with NGAL (NGAL-MMP-9) was elevated in UC patients with active disease compared to healthy controls and decreased in patients who achieved mucosal healing after infliximab therapy. These data were validated and further quantified with sandwich ELISA in the same set of samples as well as in an independent cohort of UC patients. To further validate this marker, we investigated serum NGAL-MMP-9 levels in CD patients under treatment with infliximab and found similar results (Chapter 5). In both UC and CD patients, NGAL-MMP-9 levels correlated with endoscopic and histological disease activity indices. Importantly, NGAL-MMP-9 constituted a better parameter than CRP to discriminate mucosal healing and can therefore be considered as a new biomarker for IBD.

The third aim was to investigate whether MMP-9 can be considered as a new therapeutic target in IBD. In contrast to previously reported phenotypes, we did not observe attenuation of colitis in our MMP-9 KO mice after acute DSS administration (Chapter 6). The MMP-9 KO mice did not lose less weight, had similar degree of colonic and systemic inflammation and only 11 genes involved in antimicrobial response were found to be differentially expressed compared to WT littermates. Moreover, pharmacological inhibition of MMP-9 with two peptide inhibitors (A and B) with known anti-cancer and anti-endotoxin shock properties did not improve clinical or histopathological parameters in mice with acute colitis. In fact, we observed increased Mmp3, Mmp8 and Mmp9 expression in DSS-treated mice in response to treatment with peptide inhibitor B compared to saline. Since TIMP-1 is the natural inhibitor of MMP-9 and has been shown to play a role in intestinal inflammation and fibrosis, we also evaluated the development of acute and chronic colitis in TIMP-1 KO mice (Chapter 7). After induction of acute colitis, we found that TIMP-1 KO mice had more severe colonic and systemic inflammation, whereas they recovered faster with lower weight loss and less disease activity compared to WT mice. Gene expression analysis indicated a large amount of genes that were differentially expressed between DSS-treated TIMP-1 KO and WT mice. Intriguingly, already under control conditions several immune-related genes were differentially expressed in TIMP-1 KO mice compared to WT mice, indicating a baseline sub-inflammatory status. The genetic background of the TIMP-1 KO mice as well as MMP-independent roles of TIMP-1 might play a role in these paradoxical results. After chronic DSS administration, reduced colonic inflammation and less intestinal fibrosis were observed in TIMP-1 KO mice compared to WT mice. TIMP-1 may therefore be considered as a new therapeutic target for intestinal fibrosis.

In conclusion, in this PhD project we identified that (i) the expression of many MMPs and TIMPs is dysregulated in IBD patients with active disease and is restored in patients who achieve mucosal healing, (ii) (NGAL-)MMP-9 is an excellent marker of disease activity and mucosal healing in IBD, (iii) ongoing clinical trials with MMP-9 inhibition as a therapeutic option for IBD need to be critically followed and (iv) TIMP-1 may be considered as a new therapeutic target for intestinal fibrosis.