Inflammatory bowel disease (IBD) – comprised of Crohn’s disease (CD) and ulcerative colitis (UC) – is believed to arise from a combination of genetic susceptibility and environmental factors that trigger an inappropriate mucosal immune response to constituents of the intestinal microbiome. There is now an extensive literature demonstrating that the microbiome has profound effects on immune function and, conversely, that the immune system can shape the microbiome. I hypothesized that genetic variation in mucosal immune gardening of the intestinal microbiome can result in pro-inflammatory dysbiosis, which acts as a risk factor for overt IBD. To evaluate whether individuals at risk for IBD develop dysbiosis prior to the onset of disease, a family based study was performed to characterize the microbiome and metabolome of pediatric IBD patients and their first degree relatives. These relatives are at higher risk for dysbiosis than the general population due to shared genetic and environmental factors with the IBD proband. A subset of healthy relatives in this cohort had dysbiosis with fecal metabolomic profiles (metabotypes) shared with IBD patients. The effect of the transcription factor RORgt on the intestinal microbiome was then investigated as a model of how perturbation of immune gardening could result in dysbiosis. Mice deficient in RORgt had an altered small intestinal and colonic mucosa-associated microbiome characterized by overgrowth of segmented filamentous bacteria (SFB), a microbe previously shown to promote colitis. Further knockout and cell engraftment experiments demonstrated that small intestinal gardening of SFB was mediated by RORgt-dependent T cells in a manner independent of IL-17A. The protective rs4845604 polymorphism in the RORC gene encoding RORgt was associated with altered microbial composition in mucosal wash samples from IBD patients and healthy individuals. These findings demonstrated that RORgt-dependent T cells garden the intestinal microbiome and suggest that genetic variation in this process could influence susceptibility to IBD.

It is estimated that air pollution kills 7 million people every year, and an estimated 90% of the global population live in areas with high levels of pollutants exceeding the WHO recommendations. There is a strong body of literature demonstrating the adverse effects of ambient air pollution on human health. Air pollution is a varied mix of toxic gaseous and particulate compounds, but clinical and epidemiological evidence support the particulate phase compounds as main contributors to adverse health outcomes. Our studies investigate a novel, potential mechanism linking ultrafine particles (UFPs) with intestinal inflammation. This study would be first to report not only the kinetics of microbiome effects, but also regional and longitudinal microbiome differences caused by UFP inhalation, a more physiologically relevant route of administration. We have assessed both the small intestine and colon, as well as both the mucosal and luminal microbiome. We combined controlled exposure using UFPs, the purportedly toxic component of PM, and inhalation to investigate the impact of air pollution on the gut microflora. Furthermore, our use of controlled chambers and re-aerosolization of PM selecting for UFP-range particulate matter represents significant advances in PM toxicity research. The present study leverages physiologically relevant UFP inhalation exposure, metabolic mouse models, acute and chronic IBD models, as well as microbiome bioinformatics analysis. In this dissertation, Chapter 1 discusses the basis for the need to study toxicological effects of ultrafine air particulates and lay the foundation for the relationship between air pollution, the microbiome, and inflammatory bowel disease. Chapter 2 describes microbiome effects from UFP exposure on hyperlipidemic and normolipidemic mice as well as in-vitro confirmation of UFP bioactivity. Chapter 3 characterizes the effect of UFP exposure on two acute chemically-induced mouse models of colitis. Chapter 4 discusses microbiome and inflammatory effects of UFP exposure in a genetically-modified, spontaneous chronic IBD mouse model, IL-10-/-. The work is a collaborative effort between both the Jacobs and Araujo labs at UCLA, with the assistance of the Engineering students from USC led by Dr. Constantinos Sioutas. Our studies offer insight into a potential mechanism that may explain the role of air pollution as an environmental factor contributing to rising incidence of IBD. This study is significant because we evaluate for the first time to our knowledge the effects of pulmonary UFP exposure on gut microbiome composition in IBD murine models.

Impact statement

The gut microbiome has received increasing attention as a potential modifier of disease susceptibility, but attempts to identify consistent disease-associated microbes have been met with limited success. Chapter 1 explores whether refocusing gut microbiome research away from fecal microbiomes and towards relevant intestinal regions may have more utility in defining microbes that drive disease pathology. Specifically, we find that small intestinal jejunal and ileal microbes have high predicted gut-brain and gut-metabolic functional capacity compared to microbes in the distal colon, and that these biogeographical distinctions are largely lost in mice colonized with human or mouse feces (a widely used technique in the field to establish causality). We additionally report that, while gut biogeographical patterns can be reproduced in broad strokes across facilities, ~50-60% of region-specific genera were limited to the individual facility, highlighting a central challenge in microbiome research.We next address whether disease-linked genetic risk loci can modify the gut microbiome, classifying such loci as microbiome quantitative trait loci (mb-QTL). While it is widely accepted that environmental factors affect gut microbiome composition, studies of identical twin cohorts and microbiome- genome wide association studies demonstrate that the gut microbiome is heritable to some extent. If the gut microbiome can be a vector in conveying both genetic and environmental insults to mediate disease, then this may account for the observed heterogeneity in disease incidence/susceptibility among risk loci carriers. In the 22q11.2 microdeletion mouse model of schizophrenia, we demonstrate that 22q11.2del/+ mice exhibit increased schizophrenia-like behavior compared to +/+ mice in addition to ileal but not colonic microbiome shifts (Chapter 2). In the latter half of this work, we evaluate the highly pleiotropic human A391T risk variant in the SLC39A8 gene as a potential mb-QTL in the context of two of its associated diseases - Crohn’s disease (increased risk) and Parkinson’s disease (reduced risk). We first report that SLC39A8 A393T (the murine equivalent of A391T) leads to small intestinal and colonic microbiome shifts along with inflammation in A393T mice, in a manner consistent with the effects of A393T on SLC39A8-encoded ZIP8 transport function (Chapters 3 and 4). Through transplantation of wild-type and SLC39A8 A393T microbes into antibiotic-treated wild-type mice, we show that A393T-linked microbes are sufficient to drive intestinal inflammation (Chapter 4). Lastly, through utilizing several mouse models of Parkinson’s disease (PD), we demonstrate that the microbiome- modifying A393T variant regulates susceptibility to synucleinopathy-induced PD (Chapter 5). The significance of these findings are as follows: (1) High gut-brain and gut-metabolic activity of the small intestinal microbiome warrant its consideration in experimental designs. (2) For phenotypes in which the outcomes are facility-specific, reporting gut microbiome composition alongside phenotypic outcomes is necessary and may in fact be physiologically relevant. (3) Risk loci may confer disease susceptibility through the gut microbiome. (4) Investigating microbiome-disease mechanisms of mb-QTL in preclinical studies enhances translational relevance while potentially alleviating reproducibility issues. (5) Clinically, future incorporation of microbiome information alongside genomic information could improve predictions of disease risk.

Purpose of review

Recent findings

Introduction

Methods

Results

Conclusions