Periodic Reporting for period 4 - CrUCCial (Novel diagnostic and therapeutic approach to inflammatory bowel disease based on functional characterization of patients: the CrUCCial index)
Reporting period: 2021-03-01 to 2022-08-31
Summary of the context and overall objectives of the project
Inflammatory bowel disease, or IBD, is a worldwide healthcare problem that places a considerable financial burden on both patients and the health-care system. Crohn’s disease (CD) and ulcerative colitis (UC) are at opposite extremes of the IBD continuum, which is defined by symptoms such as bloody diarrhea, abdominal cramps, perianal pain, and systemic symptoms such as weight loss. Disease is thought to emerge in genetically predisposed people who mount an inappropriate immune response to commensal microbiota and dietary antigens. Several important cellular pathways such as microbial recognition, autophagy, endoplasmic reticulum (ER) stress, and intestinal barrier function, have been identified to be involved in the pathogenesis of IBD. Not all pathways are believed to contribute in the same way, which is explaining the disease heterogeneity observed in patients. The disease's multifactorial etiology and heterogeneous nature require a personalized approach. A strategy that begins with diagnosis has the potential to help with prognosis, but it should also be applied to management. This latter point has not been addressed in IBD, where only the end outcome of chronic inflammation is addressed. The CrUCCial project has set the aim of improving the health and quality of life (QoL) of IBD patients by shedding light on disease heterogeneity and developing a novel method for disease classification. The central hypothesis of the CrUCCial was that molecular classification of the disease could eventually lead to individualized treatment strategies aimed at the specific causative defect of each patient. Therefore, we hypothesized that a functional characterization of patients for the major pathways involved in disease development would improve the management of patients and would ideally lead to a highly optimized personal regime to manage IBD, maximizing treatment response while minimizing adverse effects.
Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far
We developed and validated an ex-vivo intestinal epithelial cell culture system and showed that patients with more ER stress and autophagy risk alleles have augmented epithelial ER stress responses (Vanhove et al. JCC 2018). We generated and validated a patient-specific biopsy-derived ex vivo two-dimensional Intestinal epithelial cells (IECs) culture model. Similar to human colonic tissue slices, IEC membranes showed clear E-cadherin staining. Zonula occludens-1 localised in the apical epithelial monolayer, showing polarity. We explored the characteristic of our model further by comparing the gene expression levels of two epithelial cytokeratins over time: Cytokeratin 18, and Cytokeratin 20. Patients were then grouped according to the number of IBD-associated ER stress risk alleles (RA) in XBP1 (rs35873774) and ORMDL3 (rs2872507) or autophagy risk alleles in ATG16L1 (rs2241880), IRGM (rs10065172 and rs4958847), MTMR3 (rs2412973), LRRK2 (rs11175593) and ULK1 (rs12303764). We grouped patients into risk quartiles based on ER stress, autophagy, and their combined gene RA. Intracellular BIP, an ER stress chaperone protein, was utilised to measure ER stress with thapsigargin as ER stressor. ER stressed patient-grouped had a higher thapsigargin-mediated ER stress (BiP) induction than those with two or one RA. Using biopsy-derived IECs, we demonstrated that ER stress and autophagy genetic susceptibility may be functionally translated and evaluated in each patient.
In the primary intestinal epithelial model, we evaluated the effect of different therapies on barrier function such as butyrate which is thought to modulate intestinal barrier function positively. To assess butyrate's effect on barrier function, we created epithelial monolayers from UC patients and non-UC controls (n = 10, each). Trans-epithelial electrical resistance (TEER) was increased in epithelial monolayers treated with 8 mM butyrate without inducing apoptosis. mRNA expression of well-known epithelial barrier genes exhibited higher levels for most genes, with the highest upregulations reported for barrier-enhancing proteins CLDN1 and OCLN, while CLDN2, a pore-forming tight junction protein, was downregulated. Since inflammation is important to IBD, we incorporated TNFα and IFN as pro-inflammatory mediators in our system, reducing TEER by 14.1% compared to negative control cultures. Quantification of inflammatory proteins in apical and basolateral media of cell cultures demonstrated that butyrate does not prevent TNFα and TNF from inducing a pro-inflammatory state, but induces more inflammatory proteins than TNFα and IFN alone. The combined therapy also upregulated the apoptosis-related protein CASP8, which may explain cell morphological alterations. UC patients' TEER and barrier gene levels were compared to non-UC controls with fewer disease-associated alleles. While both had similar median TEER and barrier gene levels, there were no differences at baseline or at different treatment settings, suggesting that patients don't need higher/other butyrate concentrations or are more sensitive than healthy controls.
CruCCial index
We characterized the individual components of the CruCCial index (ECCO and UEGW 2020). The CrUCCial index was created to reflect disease pathology in each patient. The microbial dysbiosis index (MDI) was computed as the logarithm of the sum of [abundance in taxa that expanded in CD or UC] divided by [abundance in taxa that diminished in CD or UC]. Using penalised logistic regression models, we created an inflammatory proteomic score (IPS) as a weighted sum of serum levels of inflammatory proteins (OLINK proteomics). Similarly, we created a molecular score for barrier integrity, autophagy, and unfolded protein response (UPR), these were calculated as the weighted sum of relevant gene expression in the intestinal mucosa (RNA sequencing). Patients were ranked from Q1 (the least dysbiotic, inflammatory, or dysfunctional state) through Q4 for each score (the most dysbiotic, inflammatory and dysfunctional state). The CrUCCial index was calculated in a group of 312 CD patients and 148 controls. In this cohort, the MDI and IPS were both significantly associated with CRP and FC levels. None of the individual components of the score were associated with Montreal classification. In a subset of 36 patient, various microbiome components linked with the proteome. FGF-19, for example, associated positively with Faecalibacterium and negatively with Fusicatenibacterium. We found a strong positive association between MDI and IPS. We next investigated the relationship between the CrUCCial index and major clinical outcomes, such as the need for surgery or advanced biologic therapies. We furthermore showed the dynamics of the CruCCial index in patients on biological therapy, with the inflammatory proteomics score, the UPR/ER stress score and microbiome dysbiosis index being significantly decreased in responders to biologic treatment. A similar trend was observed for barrier integrity score. In 20 individuals that we had data at baseline and 6 months after initiating biological
therapy we studied the dynamics of the CrUCCial index. In this cohort the IPS and MDI decreased significantly in responders to biologic treatment (DDW 2022).
Disease heterogeneity
We investigated disease heterogeneity – a relatively understudied and therapeutically relevant aspect of IBD. In CD, disease heterogeneity includes perianal involvement, extraintestinal symptoms, disease progression, disease location and behaviour, and need for and response to diverse medications. On a molecular and cellular level, ileal, colonic, and UC differ. We integrated multi -omic datasets derived from a well-defined, heterogeneous IBD patient pool followed by systems biology approaches. By integrating gene-expression from blood derived CD4 T-cells and monocytes, we identified underlying mechanisms (eg. IL1, IL10, interferon signaling) driving critical axes of the heterogeneous disease behavior (Sudhakar et al. IBD 2021). In another study focusing on disease-location, we integrated gene expression of different blood immune cell-types from a heterogeneous IBD cohort using a wide-range of (un)supervised approaches. We identified signatures which could be relevant for location-specific interventions. For example, we identified several pathways (FOXO, MAPK) and key molecular drivers (TNFAIP3/ PTGER4/ GPR183/ NR1D2 /SIRT1/ PRKCQ) in peripheral CD4 T cells distinguishing ileal CD from UC. In addition, we identified a multi-omic biomarker panel which distinguishes patients who respond to various anti-inflammatory drugs (Verstockt et al. BMJ Open Gastro 2022). The top-drivers and key-molecular players from the heterogeneity studies were then included to furtehr enhance the clinical relevance and applicability of the CrUCCial index.
Microbiome-host interactions
The microbiome-host axis is a key fulcrum in CD, which is characterised by intestinal dysbiosis. This is reflected in IBD patients' gut microbial communities and epithelial barrier responses. Such responses cause hyperinflammation and erosion of the mucus layer, making the epithelial barrier more vulnerable to microbial influences. To get a better understanding of microbiome-host interactions (Sudhakar et al. Front. Microbiol 2021), we co-developed MicrobioLink (Andrighetti et al. Cells, 2020). To our knowledge, this integrated microbiome-host tool is one of the first to provide a mechanism-based integration of multi-omic datasets to infer inter-species cross-talk. The tool harnesses the power of protein structural features such as eukaryotic linear motifs and domains to infer the interactions between bacterial and host proteins. By using disordered regions as a quality filter, false positive predictions are minimized. Subsequently, to trace the signal flow, it uses the concept of network diffusion, which traces the path from upstream host proteins affected by the microbial proteins to the downstream host target genes/components which are differentially expressed or modulated. As a use case, we inferred mechanistic links between the microbial and host gene expression in publicly available CD patients’ datasets (Sudhakar et al. iScience, 2022). Some of the selected features were also top-ranked within the omics datasets in the CrUCCial index. These include a protease encoded within four of the eight bacterial species enriched, and was more abundant in CD patients compared to healthy controls.
In the primary intestinal epithelial model, we evaluated the effect of different therapies on barrier function such as butyrate which is thought to modulate intestinal barrier function positively. To assess butyrate's effect on barrier function, we created epithelial monolayers from UC patients and non-UC controls (n = 10, each). Trans-epithelial electrical resistance (TEER) was increased in epithelial monolayers treated with 8 mM butyrate without inducing apoptosis. mRNA expression of well-known epithelial barrier genes exhibited higher levels for most genes, with the highest upregulations reported for barrier-enhancing proteins CLDN1 and OCLN, while CLDN2, a pore-forming tight junction protein, was downregulated. Since inflammation is important to IBD, we incorporated TNFα and IFN as pro-inflammatory mediators in our system, reducing TEER by 14.1% compared to negative control cultures. Quantification of inflammatory proteins in apical and basolateral media of cell cultures demonstrated that butyrate does not prevent TNFα and TNF from inducing a pro-inflammatory state, but induces more inflammatory proteins than TNFα and IFN alone. The combined therapy also upregulated the apoptosis-related protein CASP8, which may explain cell morphological alterations. UC patients' TEER and barrier gene levels were compared to non-UC controls with fewer disease-associated alleles. While both had similar median TEER and barrier gene levels, there were no differences at baseline or at different treatment settings, suggesting that patients don't need higher/other butyrate concentrations or are more sensitive than healthy controls.
CruCCial index
We characterized the individual components of the CruCCial index (ECCO and UEGW 2020). The CrUCCial index was created to reflect disease pathology in each patient. The microbial dysbiosis index (MDI) was computed as the logarithm of the sum of [abundance in taxa that expanded in CD or UC] divided by [abundance in taxa that diminished in CD or UC]. Using penalised logistic regression models, we created an inflammatory proteomic score (IPS) as a weighted sum of serum levels of inflammatory proteins (OLINK proteomics). Similarly, we created a molecular score for barrier integrity, autophagy, and unfolded protein response (UPR), these were calculated as the weighted sum of relevant gene expression in the intestinal mucosa (RNA sequencing). Patients were ranked from Q1 (the least dysbiotic, inflammatory, or dysfunctional state) through Q4 for each score (the most dysbiotic, inflammatory and dysfunctional state). The CrUCCial index was calculated in a group of 312 CD patients and 148 controls. In this cohort, the MDI and IPS were both significantly associated with CRP and FC levels. None of the individual components of the score were associated with Montreal classification. In a subset of 36 patient, various microbiome components linked with the proteome. FGF-19, for example, associated positively with Faecalibacterium and negatively with Fusicatenibacterium. We found a strong positive association between MDI and IPS. We next investigated the relationship between the CrUCCial index and major clinical outcomes, such as the need for surgery or advanced biologic therapies. We furthermore showed the dynamics of the CruCCial index in patients on biological therapy, with the inflammatory proteomics score, the UPR/ER stress score and microbiome dysbiosis index being significantly decreased in responders to biologic treatment. A similar trend was observed for barrier integrity score. In 20 individuals that we had data at baseline and 6 months after initiating biological
therapy we studied the dynamics of the CrUCCial index. In this cohort the IPS and MDI decreased significantly in responders to biologic treatment (DDW 2022).
Disease heterogeneity
We investigated disease heterogeneity – a relatively understudied and therapeutically relevant aspect of IBD. In CD, disease heterogeneity includes perianal involvement, extraintestinal symptoms, disease progression, disease location and behaviour, and need for and response to diverse medications. On a molecular and cellular level, ileal, colonic, and UC differ. We integrated multi -omic datasets derived from a well-defined, heterogeneous IBD patient pool followed by systems biology approaches. By integrating gene-expression from blood derived CD4 T-cells and monocytes, we identified underlying mechanisms (eg. IL1, IL10, interferon signaling) driving critical axes of the heterogeneous disease behavior (Sudhakar et al. IBD 2021). In another study focusing on disease-location, we integrated gene expression of different blood immune cell-types from a heterogeneous IBD cohort using a wide-range of (un)supervised approaches. We identified signatures which could be relevant for location-specific interventions. For example, we identified several pathways (FOXO, MAPK) and key molecular drivers (TNFAIP3/ PTGER4/ GPR183/ NR1D2 /SIRT1/ PRKCQ) in peripheral CD4 T cells distinguishing ileal CD from UC. In addition, we identified a multi-omic biomarker panel which distinguishes patients who respond to various anti-inflammatory drugs (Verstockt et al. BMJ Open Gastro 2022). The top-drivers and key-molecular players from the heterogeneity studies were then included to furtehr enhance the clinical relevance and applicability of the CrUCCial index.
Microbiome-host interactions
The microbiome-host axis is a key fulcrum in CD, which is characterised by intestinal dysbiosis. This is reflected in IBD patients' gut microbial communities and epithelial barrier responses. Such responses cause hyperinflammation and erosion of the mucus layer, making the epithelial barrier more vulnerable to microbial influences. To get a better understanding of microbiome-host interactions (Sudhakar et al. Front. Microbiol 2021), we co-developed MicrobioLink (Andrighetti et al. Cells, 2020). To our knowledge, this integrated microbiome-host tool is one of the first to provide a mechanism-based integration of multi-omic datasets to infer inter-species cross-talk. The tool harnesses the power of protein structural features such as eukaryotic linear motifs and domains to infer the interactions between bacterial and host proteins. By using disordered regions as a quality filter, false positive predictions are minimized. Subsequently, to trace the signal flow, it uses the concept of network diffusion, which traces the path from upstream host proteins affected by the microbial proteins to the downstream host target genes/components which are differentially expressed or modulated. As a use case, we inferred mechanistic links between the microbial and host gene expression in publicly available CD patients’ datasets (Sudhakar et al. iScience, 2022). Some of the selected features were also top-ranked within the omics datasets in the CrUCCial index. These include a protease encoded within four of the eight bacterial species enriched, and was more abundant in CD patients compared to healthy controls.
Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)
The CruCCial index incorporates barrier integrity, autophagy, UPR/ER stress, serum inflammatory proteomics, genetic profile, and microbial dysbiosis, making it unique among clinical scoring systems. We examined its dynamic association with significant clinical outcomes and therapeutic approaches. We found the mechanisms (IL1, IL10, interferon signaling pathways, etc.) driving major axes of heterogeneity like disease behavior by integrating gene-expression from blood-derived immune cells such CD4 T-cells and monocytes. We discovered different critical pathways (FOXO, MAPK) and key molecular drivers (TNFAIP3, PTGER4, GPR183, NR1D2, SIRT1, and PRKCQ) in peripheral CD4 T cells separating ileal CD and UC patients. We also identified multi-omic biomarker panels that identify anti-inflammatory medication responders. In addition, we discovered microbe-mediated mechanisms which could be used as intervention points for microibiome-based therapeutic strategies.