Inflammatory bowel disease (IBD) is a serious, incurable gastrointestinal illness characterized by chronic inflammation which profoundly impacts quality of life for patients. Symptoms can include abdominal pain or discomfort, fever, diarrhea, and bloody stools. Up to 25% of patients experience other symptoms involving the heart, lungs, eyes, pancreas, bones, or joints. Around 1.5 million people have IBD in the U.S. and Canada, where rates are among the highest worldwide, and an estimated 17,000-93,000 new diagnoses are made each year. Importantly, peak age of onset is young adulthood, and 7-20% of diagnoses are pediatric. Because individuals require blood tests to monitor disease, surveillance colonoscopy, pharmaceutical management of symptoms, and often surgery, IBD has a large impact on the healthcare system. Recent meta-analysis of genetic data revealed 163 loci significantly associated with IBD that explain less than 15% of variance in disease risk. Intriguingly, associated loci are enriched for genes involved in host response to intestinal microbes. Previous studies have shown IBD patients' gut bacterial populations, or microbiomes, have altered configurations compared to control individuals, including overall reduction in diversity as well as altered abundance of specific bacteria. At this time, despite information that both genetic and microbiome risk factors exist for IBD, no study has taken both into account simultaneously. The validity of imputing bacterial genomes to look at microbial activity in the gut has also not been studied in IBD. We propose to investigate these issues using genotype and microbiome data from the RISK cohort, the largest collection of early-onset, treatment-nave Crohn's cases and non-IBD controls.
Our first aim i s to test the hypothesis that using genetic and microbiome data in combination is a more precise estimator of IBD risk than either genetic or microbiome profile alone. Gut microbiome dysbiosis scores will be calculated from 16S rRNA gene profiling of terminal ileum biopsy and fecal samples as described by Gevers et al. Polygenic risk scores will be calculated using significant loci and effect sizes from the recent Jostins et al. meta-analysis of IBD. We will then use logistic regression to test for association o case/control status with polygenic risk score, dysbiosis index, and both polygenic risk score and dysbiosis index.
Our second aim i s to test the hypothesis that metagenomes imputed from fecal samples more precisely estimate IBD risk compared to taxonomic information. We will impute the metagenomes of the same cohort discussed in Aim 1 using established tool PICRUSt. We can then include these gene family abundances in our model to test whether differences in the imputed metagenome demonstrate better association with case/control status compared to dysbiosis index, polygenic risk score, or the two in combination. The proposed study will be the first to evaluate IBD risk using both genetic and microbiome data in combination. Our long-term goal is to advance understanding of the mechanisms of IBD pathogenesis and help improve diagnosis and treatment. More broadly, this study will contribute important preliminary insights into how the host genome and microbiome interact.
Because genetics alone cannot predict risk for inflammatory bowel disease (IBD), we aim to improve risk assessment by combining human genetic data with information about gut bacteria. We will test whether knowledge of which bacteria are present or what genes these bacteria have can improve prediction of whether an individual has IBD. This will not only improve diagnosis but also advance our understanding of the contribution of genetic and environmental exposures to IBD pathogenesis.
|Shaw, Kelly A; Bertha, Madeline; Hofmekler, Tatyana et al. (2016) Dysbiosis, inflammation, and response to treatment: a longitudinal study of pediatric subjects with newly diagnosed inflammatory bowel disease. Genome Med 8:75|