Final Report - IBDASE (Mucosal protease and their inhibitors in inflammatory bowel disease: from etiopathogenetic insight to innovative therapy)
Executive summary:
IBDASE addresses the etiology and pathogenesis of Inflammatory bowel disease (IBD) with a multidisciplinary plan for innovative diagnosis and therapy focused on mucosal proteases and their inhibitors (P / PIs). The incidence of IBD, a multifactorial disease influenced by environmental factors in a background of complex genetic susceptibility has been rising in developed countries for the last decades. The causes for IBD are still poorly understood, although recent studies point to a crucial role of the gut microflora. There is a need for novel, more efficacious and specific therapies. In view of P / PIs as novel therapeutic targets, IBDASE evaluates their role in IBD through the combination of in vitro, in vivo and ex vivo methodologies from the diverse disciplines, such as human genetics, genetic epidemiology, biochemistry, molecular biology and immunology. The consortium identifies P / PIs that are genetically and / or functionally linked to IBD and reveals pathogenetic pathways. On the basis of these results, therapeutic approaches promoting or inhibiting mucosal P / PIs are proposed.
Project context and objectives:
1.3.1 IBD: a common disease without a cure
IBD is characterised by chronic mucosal inflammation of the gastrointestinal tract and may be complicated by extraintestinal inflammation, such as the liver or joints in the spine and pelvis. The mean age of disease onset currently peaks at less than 35 years, however the incidence has been rising in Europe over recent years, and it increasingly also affects children and infants. Nowadays, approximately 20 new IBD cases / 100 000 are diagnosed every year. The overall IBD prevalence in Europe is approximately 2 / 1000. Although current therapies counteract the inflammatory reaction in the intestinal mucosa and can temporarily induce remission and alleviate symptoms in the patient, there is no cure to the disease. Surgical interventions often are unavoidable. Therefore, IBD carries high morbidity, reduces quality of life of the affected individual and often leads to disability, leaving the patients unable to work. IBD causes causing a high burden on the health system and on the societal scale in general.
1.3.2 Etiopathogenesis in IBD
A circulus vitiosus between loss of mucosal barrier function and the intestinal microflora
Evidence derived from IBD animal models and supported by empirical observations in IBD patients indicates a crucial role for the intestinal microbiota in maintaining chronic inflammation. IBD is the consequence of a detrimental circulus vitiosus in susceptible individuals: a dysfunction of the intestinal mucosal barrier leads to adhesion and invasion of bacteria, which subsequently trigger an aberrant immune response. The inflammation this provokes is accompanied by a complete breakdown of the mucosal barrier, which further sustains the inflammatory response.
1.3.3 Objectives
The overall goal of IBDASE was the identification of the mucosal P / PI genes with highest impact on susceptibility and pathogenetic mechanisms in IBD on a genome-wide level.
Selection of initial candidate P / PI genes
Initially, the genetic evidence for an association with IBD was scored across all P / PI genes by mapping their genomic location onto the IBD susceptibility regions identified in all published genetic IBD studies in the past thirty years. We systematically reviewed the literature and generated genetic evidence scores for each P / PI gene using a proprietary algorithm. A list of candidate P / PI genes was generated based on the genetic evidence score and additional independent non-genetic information on a pathogenetic involvement in IBD.
Project results:
IBDASE addressed the role of proteases and protease inhibitors (P / PIs) in IBD, a chronic mucosal inflammation of the gastrointestinal tract, which affects 2.2 million people in Europe and 1.4 million people in North America. The activities in IBDASE were guided by the continuously updated list of candidate P / PI genes. The initial candidates on this list were selected in an unbiased fashion according to the genetic evidence for an association with Crohn's disease (CD) and Ulcerative colitis (UC) available in human genetic studies recorded in the past three decades. The initial candidate P / PI gene list was repeatedly updated and complemented with P / PI genes based on the results generated in subsequent functional studies in IBDASE. The list was finally narrowed to the most promising P / PI genes to be validated in IBDASE as therapeutical targets in IBD.
1.4.2 Genetic association - Surprising discoveries
The explorative genetic association studies of candidate P / PI genes in IBDASE were carried through in 650 CD cases, 721 UC cases and 542 control cases, and in a total of 1670 CD cases, 1391 UC cases and 1254 controls for the replication studies. As the candidate genes were identified in the systematic review based on the genetic evidence and thus selected in an unbiased fashion without any prior assumptions of a functional link to IBD, unexpected associations were found with P / PI genes involved in the Ubiquitin proteaseome system (UPS) and genes with a potential to interact with microbial components:
1.4.3 Functional studies in vitro - A promising substrate for the mucosal protease meprin alpha
Processing of mucin by meprin
Mucus is a major protein complex, establishing a physicochemical barrier towards the microbiota, which colocalises at the interface to the gut lumen with meprins, proteases expressed at high abundance by epithelial cells. The monomeric component of mucus in the intestine is the mucin 2 (MUC2), which as indicated by our data represents a promising, functionally relevant substrate for meprins. We showed that although meprin a did not affect the MUC2, meprin a is capable of cleaving both the N and C terminal parts of MUC2 at specific cleavage sites. The cleavage sites in the C terminus are close to the site for cysteine proteases from entamoeba histolytica, as described before. This region may be highly glycosylated, protecting the cleavage site. This suggests that the control of glycosylation in this region can regulate the meprin a cleavage.
Our data indicate that meprins are involved in mucin processing in vivo and meprin a is involved in the formation of the non-attached less dense mucus found in the small intestine. It has been an enigma why the mucus is different in the small (loose and non-attached) and large intestine (dense and attached) as it is built around the MUC2 in both locations. There are two models for this, namely one with an anchor protein in colon not expressed in the small intestine or the other a protease expressed in the small intestine and less in colon. The present results strongly indicate that the latter explanation is correct and that meprin a is the enzyme of the small intestine searched for. This is in line with the known distribution of meprin a with a high concentration in the small intestine and a gradual disappearance along the large intestine with no meprin a in the distal colon.
1.4.4 Functional studies in vivo - lessons from IBD animal models
To understand the regulation of P / PI gene expression, we profiled P / PI genes and immune genes of the intestine on a genome-wide level in three different IBD animal models, which reflect acute (DSS and TNBS models) and chronic intestinal inflammation (T cell transfer model), using custom-made expression microarrays. The number of regulated P / PI genes exceeds the immune genes, corroborating the important role of P / PI genes in the inflammatory response. The expression regulation in the three IBD models is very different, reflecting that different pro-inflammatory triggers in the animal models lead to unique pro-inflammatory gene signatures. In contrast to the diversity of involved genes, there are two molecular pathways that are common to the three models, namely the complement system and the UPS. Both pathways are also represented amongst the P / PI genes with greatest genetic evidence in human IBD (as evidenced by the genetic association studies in IBDASE and the genetic studies published by others), corroborating their pathogenetic role in IBD. We speculate that, similar to the IBD animal models, the triggers in human IBD and gene responses of the host are diverse. This may also be one reason for the ever increasing number of susceptibility genes in IBD, as identified in the most recent GWAS meta-analysis in CD and UC. In conclusion, therapeutic and diagnostic procedures have to be based on molecular pathways rather than IBD-associated genes alone. The modulation of the UPS and complement system harbours great therapeutic potential.
1.4.5 Functional studies: the microbiota - the protective mucosal proteolytic barrier of the host
Our objectives were to investigate pathogenetic mechanisms in IBD through analysis of proteolytic events affecting the interaction of bacteria with the epithelium and the immune system at the molecular level, to assess the role of candidate P / PIs as a component of the mucosal barrier and, vice versa, interference of bacterial derivated proteases and inhibitors with mucosal barrier functions.
The mucus layer of the colon provides a physicochemical barrier that separates the luminal microbiota from the epithelial cells, and it has been suggested that defects in this inner mucus layer should allow bacteria to reach the epithelial cells and trigger inflammation . Bacteria that secrete proteolytic enzymes are potentially able to cleave MUC2 and thereby disrupt the back-bone of the mucus layer. It has been shown that the parasite Entamoeba histolytica secretes such a mucus-degrading enzyme, which triggered the search for bacteria that might do the same.
1.4.6 Data integration - Which path does IBD go?
A major activity in IBDASE was the profiling of P / PI gene expression regulation in experimental animal models of acute and chronic intestinal inflammation (DSS, TNBS and T cell transfer). However, no pathogenetic hypothesis can be readily drawn from the list of regulated genes and candidate P / PI genes defined in IBDASE overall. We thus derived from the gene lists significantly regulated molecular pathways, which represent an analytical outcome of much higher quality. Pathway and gene ontology enrichment analysis were performed using Gene set enrichment analysis (GSEA) with human gene set annotations projected to mouse using the Mouse-genome database (MGD) ortholog assignments. Molecular pathways rather than individual genes point to pathogenetic mechanism in disease and are crucial for the selection of therapeutic targets.
Significantly regulated molecular pathways: the complement pathway and the UPS
1.4.7 Innovative therapy - reinforcing the proteolytic barrier
We validated selected candidate P / PI genes as therapeutic targets, modulating their activity in targeted IBD animal models. For instance, a pilot experiment with actinonin, an inhibitor for meprin, in the Trinitrobenzene sulfonic acid (TNBS) model, showed a worse outcome in the presence of the meprin inhibitor, consistent with a protective role of meprin in the intestinal mucosa. The same observation was made with a range of different mucosal P / PIs. These findings support our concept of the mucosal protection by the constitutive proteolytic barrier. Our data thus indicate that mucosal proteolytic activity, including meprin, must be restored rather than inhibited in inflamed intestine.
In this context, enhanced wound healing is expected to support restoration of the proteolytic barrier, which is mainly contributed by the intact epithelial cell layer. We observed in an in vitro wound healing assay using human intestinal fibroblasts that small interfering ribonucleic acid (siRNA)-mediated downregulation of certain proteases enhanced cell migration and wound healing, especially in the presence of TNF-alpha, which is abundant in inflammatory conditions in IBD. SiRNA-mediated knockdown of selected P / PI genes could therefore be used as a therapeutic approach to boost wound healing in inflammatory diseases, such as in IBD. However, in vivo delivery of siRNA remains a substantial hurdle to overcome. Although siRNA was shown to be effective at the nanometre range in vitro, it but could not be delivered to the target site in the intestine in vivo. More work is needed to translate the promises of siRNA-mediated therapy to the local in vivo environment of IBD.
List of websites: http://ibdase.org
IBDASE addresses the etiology and pathogenesis of Inflammatory bowel disease (IBD) with a multidisciplinary plan for innovative diagnosis and therapy focused on mucosal proteases and their inhibitors (P / PIs). The incidence of IBD, a multifactorial disease influenced by environmental factors in a background of complex genetic susceptibility has been rising in developed countries for the last decades. The causes for IBD are still poorly understood, although recent studies point to a crucial role of the gut microflora. There is a need for novel, more efficacious and specific therapies. In view of P / PIs as novel therapeutic targets, IBDASE evaluates their role in IBD through the combination of in vitro, in vivo and ex vivo methodologies from the diverse disciplines, such as human genetics, genetic epidemiology, biochemistry, molecular biology and immunology. The consortium identifies P / PIs that are genetically and / or functionally linked to IBD and reveals pathogenetic pathways. On the basis of these results, therapeutic approaches promoting or inhibiting mucosal P / PIs are proposed.
Project context and objectives:
1.3.1 IBD: a common disease without a cure
IBD is characterised by chronic mucosal inflammation of the gastrointestinal tract and may be complicated by extraintestinal inflammation, such as the liver or joints in the spine and pelvis. The mean age of disease onset currently peaks at less than 35 years, however the incidence has been rising in Europe over recent years, and it increasingly also affects children and infants. Nowadays, approximately 20 new IBD cases / 100 000 are diagnosed every year. The overall IBD prevalence in Europe is approximately 2 / 1000. Although current therapies counteract the inflammatory reaction in the intestinal mucosa and can temporarily induce remission and alleviate symptoms in the patient, there is no cure to the disease. Surgical interventions often are unavoidable. Therefore, IBD carries high morbidity, reduces quality of life of the affected individual and often leads to disability, leaving the patients unable to work. IBD causes causing a high burden on the health system and on the societal scale in general.
1.3.2 Etiopathogenesis in IBD
A circulus vitiosus between loss of mucosal barrier function and the intestinal microflora
Evidence derived from IBD animal models and supported by empirical observations in IBD patients indicates a crucial role for the intestinal microbiota in maintaining chronic inflammation. IBD is the consequence of a detrimental circulus vitiosus in susceptible individuals: a dysfunction of the intestinal mucosal barrier leads to adhesion and invasion of bacteria, which subsequently trigger an aberrant immune response. The inflammation this provokes is accompanied by a complete breakdown of the mucosal barrier, which further sustains the inflammatory response.
1.3.3 Objectives
The overall goal of IBDASE was the identification of the mucosal P / PI genes with highest impact on susceptibility and pathogenetic mechanisms in IBD on a genome-wide level.
Selection of initial candidate P / PI genes
Initially, the genetic evidence for an association with IBD was scored across all P / PI genes by mapping their genomic location onto the IBD susceptibility regions identified in all published genetic IBD studies in the past thirty years. We systematically reviewed the literature and generated genetic evidence scores for each P / PI gene using a proprietary algorithm. A list of candidate P / PI genes was generated based on the genetic evidence score and additional independent non-genetic information on a pathogenetic involvement in IBD.
Project results:
IBDASE addressed the role of proteases and protease inhibitors (P / PIs) in IBD, a chronic mucosal inflammation of the gastrointestinal tract, which affects 2.2 million people in Europe and 1.4 million people in North America. The activities in IBDASE were guided by the continuously updated list of candidate P / PI genes. The initial candidates on this list were selected in an unbiased fashion according to the genetic evidence for an association with Crohn's disease (CD) and Ulcerative colitis (UC) available in human genetic studies recorded in the past three decades. The initial candidate P / PI gene list was repeatedly updated and complemented with P / PI genes based on the results generated in subsequent functional studies in IBDASE. The list was finally narrowed to the most promising P / PI genes to be validated in IBDASE as therapeutical targets in IBD.
1.4.2 Genetic association - Surprising discoveries
The explorative genetic association studies of candidate P / PI genes in IBDASE were carried through in 650 CD cases, 721 UC cases and 542 control cases, and in a total of 1670 CD cases, 1391 UC cases and 1254 controls for the replication studies. As the candidate genes were identified in the systematic review based on the genetic evidence and thus selected in an unbiased fashion without any prior assumptions of a functional link to IBD, unexpected associations were found with P / PI genes involved in the Ubiquitin proteaseome system (UPS) and genes with a potential to interact with microbial components:
1.4.3 Functional studies in vitro - A promising substrate for the mucosal protease meprin alpha
Processing of mucin by meprin
Mucus is a major protein complex, establishing a physicochemical barrier towards the microbiota, which colocalises at the interface to the gut lumen with meprins, proteases expressed at high abundance by epithelial cells. The monomeric component of mucus in the intestine is the mucin 2 (MUC2), which as indicated by our data represents a promising, functionally relevant substrate for meprins. We showed that although meprin a did not affect the MUC2, meprin a is capable of cleaving both the N and C terminal parts of MUC2 at specific cleavage sites. The cleavage sites in the C terminus are close to the site for cysteine proteases from entamoeba histolytica, as described before. This region may be highly glycosylated, protecting the cleavage site. This suggests that the control of glycosylation in this region can regulate the meprin a cleavage.
Our data indicate that meprins are involved in mucin processing in vivo and meprin a is involved in the formation of the non-attached less dense mucus found in the small intestine. It has been an enigma why the mucus is different in the small (loose and non-attached) and large intestine (dense and attached) as it is built around the MUC2 in both locations. There are two models for this, namely one with an anchor protein in colon not expressed in the small intestine or the other a protease expressed in the small intestine and less in colon. The present results strongly indicate that the latter explanation is correct and that meprin a is the enzyme of the small intestine searched for. This is in line with the known distribution of meprin a with a high concentration in the small intestine and a gradual disappearance along the large intestine with no meprin a in the distal colon.
1.4.4 Functional studies in vivo - lessons from IBD animal models
To understand the regulation of P / PI gene expression, we profiled P / PI genes and immune genes of the intestine on a genome-wide level in three different IBD animal models, which reflect acute (DSS and TNBS models) and chronic intestinal inflammation (T cell transfer model), using custom-made expression microarrays. The number of regulated P / PI genes exceeds the immune genes, corroborating the important role of P / PI genes in the inflammatory response. The expression regulation in the three IBD models is very different, reflecting that different pro-inflammatory triggers in the animal models lead to unique pro-inflammatory gene signatures. In contrast to the diversity of involved genes, there are two molecular pathways that are common to the three models, namely the complement system and the UPS. Both pathways are also represented amongst the P / PI genes with greatest genetic evidence in human IBD (as evidenced by the genetic association studies in IBDASE and the genetic studies published by others), corroborating their pathogenetic role in IBD. We speculate that, similar to the IBD animal models, the triggers in human IBD and gene responses of the host are diverse. This may also be one reason for the ever increasing number of susceptibility genes in IBD, as identified in the most recent GWAS meta-analysis in CD and UC. In conclusion, therapeutic and diagnostic procedures have to be based on molecular pathways rather than IBD-associated genes alone. The modulation of the UPS and complement system harbours great therapeutic potential.
1.4.5 Functional studies: the microbiota - the protective mucosal proteolytic barrier of the host
Our objectives were to investigate pathogenetic mechanisms in IBD through analysis of proteolytic events affecting the interaction of bacteria with the epithelium and the immune system at the molecular level, to assess the role of candidate P / PIs as a component of the mucosal barrier and, vice versa, interference of bacterial derivated proteases and inhibitors with mucosal barrier functions.
The mucus layer of the colon provides a physicochemical barrier that separates the luminal microbiota from the epithelial cells, and it has been suggested that defects in this inner mucus layer should allow bacteria to reach the epithelial cells and trigger inflammation . Bacteria that secrete proteolytic enzymes are potentially able to cleave MUC2 and thereby disrupt the back-bone of the mucus layer. It has been shown that the parasite Entamoeba histolytica secretes such a mucus-degrading enzyme, which triggered the search for bacteria that might do the same.
1.4.6 Data integration - Which path does IBD go?
A major activity in IBDASE was the profiling of P / PI gene expression regulation in experimental animal models of acute and chronic intestinal inflammation (DSS, TNBS and T cell transfer). However, no pathogenetic hypothesis can be readily drawn from the list of regulated genes and candidate P / PI genes defined in IBDASE overall. We thus derived from the gene lists significantly regulated molecular pathways, which represent an analytical outcome of much higher quality. Pathway and gene ontology enrichment analysis were performed using Gene set enrichment analysis (GSEA) with human gene set annotations projected to mouse using the Mouse-genome database (MGD) ortholog assignments. Molecular pathways rather than individual genes point to pathogenetic mechanism in disease and are crucial for the selection of therapeutic targets.
Significantly regulated molecular pathways: the complement pathway and the UPS
1.4.7 Innovative therapy - reinforcing the proteolytic barrier
We validated selected candidate P / PI genes as therapeutic targets, modulating their activity in targeted IBD animal models. For instance, a pilot experiment with actinonin, an inhibitor for meprin, in the Trinitrobenzene sulfonic acid (TNBS) model, showed a worse outcome in the presence of the meprin inhibitor, consistent with a protective role of meprin in the intestinal mucosa. The same observation was made with a range of different mucosal P / PIs. These findings support our concept of the mucosal protection by the constitutive proteolytic barrier. Our data thus indicate that mucosal proteolytic activity, including meprin, must be restored rather than inhibited in inflamed intestine.
In this context, enhanced wound healing is expected to support restoration of the proteolytic barrier, which is mainly contributed by the intact epithelial cell layer. We observed in an in vitro wound healing assay using human intestinal fibroblasts that small interfering ribonucleic acid (siRNA)-mediated downregulation of certain proteases enhanced cell migration and wound healing, especially in the presence of TNF-alpha, which is abundant in inflammatory conditions in IBD. SiRNA-mediated knockdown of selected P / PI genes could therefore be used as a therapeutic approach to boost wound healing in inflammatory diseases, such as in IBD. However, in vivo delivery of siRNA remains a substantial hurdle to overcome. Although siRNA was shown to be effective at the nanometre range in vitro, it but could not be delivered to the target site in the intestine in vivo. More work is needed to translate the promises of siRNA-mediated therapy to the local in vivo environment of IBD.
List of websites: http://ibdase.org