Tim Elliott

Genome-wide association study of ulcerative colitis identifies three new susceptibility loci, including the HNF4A region

Ulcerative colitis is a common form of inflammatory bowel disease with a complex etiology. As part of the Wellcome Trust Case Control Consortium 2, we performed a genome-wide association scan for ulcerative colitis in 2,361 cases and 5,417 controls. Loci showing evidence of association at P < 1 × 10−5 were followed up by genotyping in an independent set of 2,321 cases and 4,818 controls. We find genome-wide significant evidence of association at three new loci, each containing at least one biologically relevant candidate gene, on chromosomes 20q13 (HNF4A; P = 3.2 × 10−17), 16q22 (CDH1 and CDH3; P = 2.8 × 10−8) and 7q31 (LAMB1; P = 3.0 × 10−8). Of note, CDH1 has recently been associated with susceptibility to colorectal cancer, an established complication of longstanding ulcerative colitis. The new associations suggest that changes in the integrity of the intestinal epithelial barrier may contribute to the pathogenesis of ulcerative colitis.

Distinguishing orofacial granulomatosis from crohn's disease: Two separate disease entities?

Background: Orofacial granulomatosis (OFG) is a rare chronic inflammatory disease of unknown etiology sharing histological features with Crohns disease (CD). This study aimed to 1) define the clinical presentation of OFG, 2) establish differentiating features for those with CD, 3) examine if onset of OFG is predictive of CD, and 4) establish differentiating features for children. Methods: Data were extracted from medical notes (n = 207) for demographics, clinical features, blood parameters, diagnosis of CD, and treatments for patients with OFG. Results: Ninety‐seven patients (47%) were female. The lips (184/203; 91%) and buccal mucosa (151/203; 74%) were mainly affected. Forty‐six (22%) had intestinal CD. Ulcers (24/46; 46% versus 29/159; 15%, P = <0.001) were more common in patients with CD as was a raised C‐reactive protein (24/33; 73% versus 60/122; 49%, P = 0.016) and abnormal full blood count (19/41; 46% versus 35/150; 23%). The buccal‐sulcus (12/44; 27% versus 20/158; 13%, P = 0.019) was more often affected in those with CD. Half the patients with CD were diagnosed prior to onset of OFG. The remainder were diagnosed after. The incidence of CD is similar for children (16/69; 23%) and adults (29/132; 22%), although oral onset in childhood is more likely to occur prior to diagnosis of CD. Conclusions: OFG mainly presents in young adults with lip and buccal involvement. Abnormalities in inflammatory markers, hematology and oral features of ulceration, and buccal‐sulcal involvement are factors more commonly associated with CD. Initial presentation of OFG does not necessarily predict development of CD, although this is more likely in childhood. (Inflamm Bowel Dis 2011;)

Further experience with the use of 6-thioguanine in patients with Crohn's disease

Background: 6‐Thioguanine (6‐TG) is efficacious in patients with Crohns Disease (CD) failing conventional immunosuppression but there are reports of hepatotoxicity. We report our experience of the safety and efficacy of 6‐TG in a series of patients with CD. Methods: A retrospective study of patients with CD who failed thiopurines ± methotrexate between 2001 and 2006 was performed. Indications for 6‐TG were; active disease, to allow infliximab withdrawal, steroid sparing, or fistula closure. Patients underwent regular review and those treated longer than 1 year were advised to have liver magnetic resonance imaging (MRI) and liver biopsy. Results: All 30 patients treated with 6‐TG during the period were included. The median dose and duration of 6‐TG was 40 mg daily and 21.5 months, respectively. Initial clinical response was achieved in 18/30 (60%). Eleven of 29 (38%) (1 unrelated death) remained in remission at a median 44 months follow‐up. Seven of 30 (23%) discontinued 6‐TG due to adverse effects; 7/30 (23%) patients developed abnormal liver function tests (LFTs) during treatment, mostly transient and mild. One patient developed a portal hypertensive syndrome resolving on cessation of 6‐TG. Of 11 liver biopsies, none showed nodular regenerative hyperplasia (NRH). The median red blood cell 6‐thioguanine nucleotide (6‐TGN) level was 807 pmol/108. Conclusions: 6‐TG has good clinical efficacy for third‐line immunosuppression in CD but hepatotoxicity remains a concern. However, previous reports of NRH in 6‐TG‐treated inflammatory bowel disease patients have not been substantiated by this cohort.

Quantification and characterization of mucosa-associated and intracellular Escherichia coli in inflammatory bowel disease.

Background:Mucosa-associated Escherichia coli are abundant in inflammatory bowel disease (IBD), but whether these bacteria gain intracellular access within the mucosa is uncertain. If E. coli does gain intracellular access, the contribution of bacterial pathogenicity to this requires further elucidation. This study aimed to quantify and characterize mucosa-associated and intracellular E. coli in patients with IBD and in healthy control subjects (HC). Methods:Mucosal biopsies from 30 patients with Crohns disease (CD), 15 with ulcerative colitis (UC), and 14 HC were cultured with or without gentamicin protection to recover intracellular or mucosa-associated E. coli, respectively. Overall, 40 strains (CD: n = 24, UC: n = 9, and HC: n = 7) were characterized by phylogenetic typing, adhesion and invasion assays, detection of virulence factors, antimicrobial resistance genes, and proteomic analysis. Results:Mucosa-associated E. coli were more abundant in CD and UC than in HC (2750 versus 1350 versus 230 median colony-forming units per biopsy; P = 0.01). Intracellular E. coli were more prevalent in CD (90%) than in UC (47%) or HC mucosal biopsies (0%) (P < 0.001). Of 24 CD strains, 2 were adherent and invasive, but there were no unifying pathogenicity determinants that could distinguish most CD strains from UC or HC strains, or intracellular isolates from mucosa-associated isolates. Conclusions:Intracellular E. coli are more common in CD than in UC and not identified in HC. Most intracellular E. coli did not have characterizing pathogenic features, suggesting a significant role for defects in mucosal immunity or barrier dysfunction in their ability to gain intracellular access.

Defective macrophage handling of Escherichia coli in Crohn's disease

 Escherichia coli can be isolated from lamina propria macrophages in Crohns disease (CD), and their intramacrophage persistence may provide a stimulus for inflammation. To further determine the contributions of macrophage dysfunction and E. coli pathogenicity to this, we aimed to compare in vitro functioning of macrophages from patients with CD and healthy controls (HC) in response to infection with CD‐derived adherent‐invasive E. coli (AIEC) and less pathogenic E. coli strains.

Genetic Association Analysis Reveals Differences in the Contribution of NOD2 Variants to the Clinical Phenotypes of Orofacial Granulomatosis.

Background:Orofacial granulomatosis (OFG) is a rare, inflammatory disorder of the mouth, in which some patients also have intestinal Crohns disease (CD). The etiology remains largely unknown, although there is a high prevalence of atopy, and oral granulomas are also seen in other immune disorders particularly CD and sarcoidosis. We investigated whether genetic variants associated with an increased risk of CD, sarcoidosis, or atopy were also associated with susceptibility to OFG. Methods:Patients were stratified clinically as isolated oral manifestations (OFG only) or concurrent intestinal CD (OFG+CD). We genotyped 201 patients and 1023 healthy controls for risk variants in NOD2, IRGM, IL23R, ATG16L1 (CD), BTNL2 (sarcoidosis), and FLG (atopy). The coding regions of the NOD2 gene were screened for rare, potentially pathogenic variants in OFG. Results:A combined analysis of 3 CD-risk variants in NOD2 showed no association with any OFG subgroup. NOD2 p.L1007insC was associated with OFG+CD (P = 0.023) and IL23R p.R381Q with all OFG (P = 0.031). The sarcoidosis risk variant rs2076530 in BTNL2 was associated with all OFG (P = 0.013). We identified 7 rare missense NOD2 alleles in 8 individuals with OFG, 4 OFG-only patients and 4 patients with OFG+CD. There was a significant enrichment of NOD2 variants in the OFG+CD group compared to the OFG-only group (P = 0.008, common variants; P = 0.04, all common and rare variants). Conclusions:Our findings suggest that genetic variants in NOD2 are only associated with OFG in patients with concurrent intestinal disease. A genome-wide association scan is needed to fully define the genetic architecture of OFG.

Defective Macrophage Function in Crohn's Disease: Role of Alternatively Activated Macrophages in Inflammation

Introduction The aetiology of Crohns disease (CD) involves a genetically determined dysregulated immune response to commensal intestinal microflora. In CD, viable E coli are found within lamina propria macrophages (MΦs) and E coli intracellular survival is prolonged in CD-derived MΦs in vitro. Different MΦ subpopulations exist, M1 cells are inflammatory cells, M2a cells are involved in tissue repair and M2c are regulatory cells that limit inflammation. The role these different MΦs play in this abnormal handling of bacteria in CD is unclear. Methods The aim of this study was to examine in vitro M1 and M2 MΦ maturation in CD patients and healthy individuals and how these cells respond to challenge with CD-derived E coli. To do this we monitored MΦ morphology, surface markers and cytokine production and intracellular bacterial survival. Peripheral blood monocytes were isolated from CD patients and healthy controls and treated with cytokines to generate distinct MΦ subpopulations: IFNγ for M1, IL4/IL13 for M2a and IL10 for M2c. MΦ morphology was assessed by H&E staining and surface marker expression of CD14, CD16 and CD33, chemokine receptor CCR2, scavenger receptor CD163, co stimulatory molecule CD40 and mannose receptor CD206 using flow cytometry. IL10 and TNFα production were measured by ELISA and intracellular E coli survival was measured using the gentamicin protection assay. Results In contrast to M1 MΦs, both CD-derived M2 populations failed to develop the characteristic MΦ dendrite morphology seen in control macrophages. Surface expression of CD40 in M2 CD-derived MΦs was 3.4-fold lower for M2a and 4.4-fold lower for M2c compared to controls after E coli challenge. CD163 was higher in M1 CD cells but reduced by 50% in M2 cells compared to healthy cells. After E coli challenge, TNFα production by M2 but not M1 MΦs was significantly lower in CD than in controls (M2a 38%, M2c 27% respectively less than controls) but there were no differences in IL10 production. Prolonged intracellular survival of E coli was demonstrated in CD M2 cells but effective killing occurred in all M1 CD MΦs and all control MΦs. Conclusion In CD, M2 (but not M1) MΦs display abnormal morphological maturation, lower TNFα levels after E coli challenge, prolonged intracellular bacterial survival and differences in surface marker expression. The results are consistent with an innate MΦ defect in CD relating particularly to a failure of the normal role of M2 MΦs to limit and control inflammation.

Autophagy Gene Polymorphisms Influence the Interaction of E.Coli and Macrophages in Crohn's Disease

Introduction A well established expression of the proposed defect in innate immunity in Crohns disease (CD) is the prolonged survival of strains of Escherichia coli in peripheral blood monocytes/macrophages (PBM) from patients with CD compared to those from healthy controls (HC). Altered handling of E coli in CD has been demonstrated in PBM and in-situ in CD by laser dissection of E coli infected macrophages. Several of the gene polymorphisms conferring an increased risk of CD occur in genes involved in bacteria-immune cell interactions. The aim of this study was to determine the influence of these polymorphisms on the interaction of E coli with CD-derived monocyte / macrophages. Methods PBMs and lamina propria macrophages were obtained from 43 and 17 CD patients respectively (mean age 39 years / range 20–70 years). PBMs were challenged with E coli and then the response of these PBMs measured by levels of bacterial killing, using the gentamicin protection assay. Phagocytosis and reactive oxygen radicals (RORs) production were measured by flow cytometry using the phagotest and phagoburst assays. Cytokine (TNFα, IL10, IL23) production was measured by ELISA. In parallel experiments, TNFα, IL10 and CD163 expression was measured by qRT-PCR in laser dissected E coli laden and unladen lamina proprial macrophages. All patients were genotyped for the three common NOD-2 SNPs (G908R, L1007fs and R702W), and the autophagy related SNPs (ATG16L1 and IRGM). Genotypes and monocyte/macrophage markers were compared. Results Presence of the ATG16L1 risk variant was strongly associated with a reduced E coli induced TNFα expression in CD PBMs (p=0.0017) and increased levels of RORs on uptake of CD derived E coli (p=0.0027). Presences of IRGM risk variant was associated with increased IL10 expression from PBMs. NOD-2 mutations were not associated with a difference in any variable studied. No genotype altered survival of E coli in PBM at 2 or 4 h after infection or in-situ lamina propria macrophage cytokine expression. Conclusion These results show that defective autophagy modifies the handling of intracellular bacteria in CD but appears unlikely to be the main cause of prolonged bacterial survival. Further studies are needed to elucidate the nature of this innate defect and its role in CD pathogenesis.

Virulence Factors, Antimicrobial Resistance Genes and Protein Structure of Crohn's Disease-Derived E.Coli Strains Identified Using DNA Microarray and MALDI-TOF Analysis

Introduction: Intracellular E. coli isolates are found in up to 60% of gut biopsies from patients with Crohns disease (CD). Although adherent invasive E.coli (AIEC) are described[1], other intracellular E. coli are frequently isolated but specific pathogenic determinants for these have not been identified by conventional techniques. DNA microarray and Matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) mass spectrometry enable rapid and accurate DNA and proteomic analysis of bacteria. Aims: To investigate for virulence and antimicrobial resistance (AMR) genes in, and analyse the structure of CD-related intracellular E.coli isolates. Methods: Seventeen E.coli strains were isolated using gentamicin protection assay from gut biopsies of 17 CD patients (ileal n=7, colonic n=10). DNA was extracted for analysis by DNA microarray of 392 virulence and AMR genes. DNA PCR was performed for an additional 18 virulence genes. For MALDI-TOF sample preparation, several colonies of each bacterial strain were spotted onto target plate and prepared as described for Bruker Biotyper MALDI TOF analysis. Ten replicates of the mass/charge spectra for each strain were compared to spectra from a bacterial database Results: Virulence factors were identified by DNA microarray in a number of E.coli strains including senB (encodes enterotoxin), Iha (encodes an adhesin) Sat (toxin of uropathogenic E.Coli), iron regulation genes and microcin synthesis genes. DNA PCR revealed additional virulence factors including ibeA (associated with E.coli invasion in neonatal meningitis) in 2 strains. A minority of strains were positive for AMR genes relating to beta-lactams, sulphonamides, tetracycline and streptomycin. MALDI-TOF analysis confirmed the bacteria as E.coli and clustered the Crohns disease E.coli strains together and separate from other E.coli strains according to protein structure on the Bruker database. There was no clustering within CD strains in relation to site of disease or biopsy, or use of immunomodulation. Conclusion: Although CD appears to be at least in part a host immune disorder, pathogenic factors of various organisms may still play a role in pathogenesis. A number of genes encoding for virulence factors have been identified in a proportion of CD-related E.coli strains which warrant further investigation (eg whole genome sequencing and functional assays). Most strains had no antimicrobial resistance genes. Further MALDI-TOF comparison of CD-related bacterial structure with commensal and IBD-related bacteria may be revealing. References 1. Miquel S et al. PLoS One. 2010 Sep 17;5(9). pii: e12714

Immune profile of gut mucosa t cells associated with e coli laden macrophages in crohn's disease

Introduction There is increasing evidence that a defect in the handling of intracellular bacteria is involved in the pathogenesis of Crohns disease (CD). In CD, we have previously demonstrated a population of intestinal macrophages harbouring viable E coli and it is therefore possible that this failure of bacterial killing contributes to the observed inflammatory response. Previous studies showed that E coli laden (E coli positive) macrophages produce less TNFα compared to those not infected with bacteria (E coli negative). T cells are seen clustering around these infected macrophages suggesting T cell activation is altered between the innate and adaptive immune systems. Aims To use co-localisation immunofluorescent techniques to investigate the immunological profile of T cells associated with E coli positive and negative macrophages, and those from control mucosa. Methods Snap frozen mucosal biopsies were taken at routine colonoscopy from patients with CD and controls with normal colorectal mucosa and cryostat section made. E coli positive/negative macrophages were identified using CD-68 staining and a polyclonal anti E coli antibody and the T cells were identified using an anti CD3 antibody. In serial sections the T cells were labelled with anti IL10, IL17, TNFα, TGFβ, FoxP3 or IL23 receptor and the results analysed using confocal microscopy. Results The profile of T cells surrounding E coli positive macrophages compared to E coli negative macrophages demonstrated an elevated number of cells with a regulatory phenotype. More of these T cells expressed IL-10 (67% vs 11%; p=0.001) and FoxP3 (14% vs 2%; p= 0.001). In contrast, increased numbers of pro-inflammatory T cells were seen surrounding E coli negative macrophages as measured by cells staining positive for TNFα (22%), IL-17 (26%) and IL23 (30%), which were not seen in association with E coli positive macrophages. Conclusion There is an increased number and specific distribution of regulatory or inflammatory T cells surrounding lamina propria macrophages in CD according to intracellular persistence of E coli. The results suggest that despite E coli persistence, an appropriate immune containment occurs while it is the macrophages and T cells free of E coli which make the main contribution to active inflammation. Further studies are needed to determine the immune interplay between these populations and whether other microbial factors are present in E coli negative macrophages.