Final Progress Report
Proposal No. IBD-0246
Principal Investigator: Herbert W. Virgin, M.D., Ph.D.
Applicant Organization: Washington University (St. Louis, Missouri, U.S.A.)
Project Title: In vivo function of Crohn’s disease susceptibility gene ATG16L1 in intestinal inflammation
Period of Award: September 1, 2008 – February 28, 2011
Summary of project aims:
A polymorphism in the autophagy gene ATG16L1 was recently identified as a susceptibility allele for Crohn’s Disease. It is unclear how ATG16L1 contributes to intestinal disease. Moreover, similar to other genetic variations associated with Crohn’s disease, this polymorphism in ATG16L1 is common. Why this and other polymorphisms are associated with Crohn’s disease in only a subset of individuals is unknown. To address these fundamental questions, we generated novel mouse models with mutations in Atg16L1. These mice are referred to as Atg16L1HM (HM=hypomorphic). We found that Atg16L1HM mice have striking abnormalities in Paneth cells, specialized intestinal epithelial cells with important roles in innate immunity. Importantly, this observation in mice allowed us to identify a Paneth cell abnormality in human Crohn’s disease patients that are homozygous for the risk allele of ATG16L1. Interestingly, we observed that Paneth cell abnormalities in Atg16L1HM mice were dependent on the mouse barrier facility in which they were housed. Based on these findings, we previously proposed the following specific aims:
1. Test the hypothesis that mammalian Atg16L1 is required for autophagy.
2. Test the hypothesis that intestinal pathology observed in Atg16L1HM mice is associated with a specific enteric pathogen.
3. Test the hypothesis that the pathology associated with Atg16L1-deficiency is due to an autophagy defect in the intestinal epithelium.
4. Test the hypothesis that CD patients carrying the susceptibility allele have increased Paneth cell defects.
Accomplishments toward meeting these aims:
We have made significant progress towards testing the above hypotheses, and as a result, have significantly contributed to our understanding of how ATG16L1 function determines disease susceptibility. In addition to addressing the above aims, we have also identified a novel intestinal injury response in the Atg16L1HM mice. This injury response is highly relevant to Crohn’s disease (see below). Remarkably, we have also discovered a novel role for a viral infection in triggering intestinal abnormalities in Atg16L1HM mice. Both Paneth cell abnormalities and the aberrant intestinal injury response are dependent on infection with the murine norovirus (MNV). Since MNV triggers intestinal disease specifically in Atg16L1HM and not control mice, our results suggest that the ability of an infectious agent to act as a trigger may be dependent on an individual's genetic background. We believe that this disease model will set the standard for investigating Crohn's disease etiology and guide mechanistic studies. The paper describing these findings has been accepted into Cell and will be published in June of 2010.
List of significant results:
1. Atg16L1 is required for conventional autophagy. Mammalian Atg16L1 was previously not characterized, and we confirmed that it is a bona fide autophagy protein.
2. Atg16L1HM mice display additional Paneth cell abnormalities. In addition to aberrant granule morphology, we found that Atg16L1 mutant Paneth cells were unable package and exocytosis antimicrobial granules. Moreover, Atg16L1 mutant Paneth cells displayed severe structural defects at the organelle level and had an inflammatory transcriptional profile.
3. Crohn's disease patients with the risk allele of ATG16L1 have additional Paneth cell abnormalities. Initially, we detected Paneth cell granule abnormalities by histology in 100% of the patients homozygous for the risk allele of ATG16L1 compared to none of the patients homozygous for the non-risk allele. Since we detected further abnormalities in Atg16L1HM mice, we investigated if Crohn's disease patients with the risk allele also shared these abnormalities. Similar to Atg16L1HM mice, we found that only patients homozygous for the risk allele of ATG16L1 had abnormal distribution of the granule protein lysozyme and overexpressed the inflammatory molecule leptin. Taken together, our results demonstrate a remarkable concordance between mice and humans with mutations in Atg16L1.
4. Mice with conditional deletion of either Atg5 or Atg7 in the intestinal epithelium display similar Paneth cell abnormalities as Atg16L1HM mice. Since Atg5 and Atg7 are autophagy proteins, this result further associates the autophagy pathway with Paneth cell function. Future experiments must address if Paneth cell abnormalities are due to a defect in conventional autophagy or an as yet unidentified role for this part of the autophagy pathway.
5. Atg16L1HM mice develop Crohn’s disease-like pathology in response to intestinal injury. When challenged with the intestinal damaging agent dextran sodium sulfate (DSS), Atg16L1HM mice develop several pathologies that are frequently described in Crohn’s disease patients. This aberrant response to DSS is dependent on TNFα, IFNγ, and commensal bacteria – three factors that are associated with human Crohn’s disease.
6. Intestinal abnormalities in Atg16L1HM mice are triggered by viral infection. Both Paneth cell abnormalities and the aberrant injury response to DSS require both Atg16L1 mutation and MNV infection, neither of which are sufficient on their own.
7. Intestinal abnormalities in Atg16L1HM mice are dependent on the strain of the virus and the timing of the infection. Conventional techniques often do not distinguish between closely related strains of a virus and cannot discern when an individual was infected. If human disease is also dependent on the strain and timing of the viral infection, then these conventional techniques may overlook potential etiological agents.
Significance of these findings. Consistent with the heterogeneity observed across Crohn's disease patients, our results suggest that disease etiology and mechanism may be dependent on an individual's genetic background. Understanding the mechanism by which various polymorphisms increase or decrease susceptibility to an infectious trigger will help guide personalized healthcare and disease prevention strategies.
Lay summary of the progress report:
The intestinal inflammatory condition known as Crohn's disease is complex. Instead of one mutation that causes the disease, there are many genetic variations that are associated with increased risk. There is also accumulating evidence suggesting that infectious agents trigger Crohn's disease, but this remains a controversial topic. One factor recently associated with Crohn's disease is a variation in the gene ATG16L1. Interestingly, this 'risk' variant of ATG16L1 is found in many healthy individuals, not just Crohn's disease patients. Therefore, we wanted to address two major questions. What does ATG16L1 have to do with inflammation in the intestine? Also, why do only some individuals with the risk variant of ATG16L1 develop Crohn's disease?
To address these questions, we generated mice with mutations in ATG16L1. We found that ATG16L1 mutant mice develop several dramatic pathologies in the intestine. Many of these inflammatory pathologies are also found in Crohn's disease patients. One of these pathologies was particularly interesting because it had not been previously described. We identified striking abnormalities in a specialized intestinal cell-type called Paneth cells. Paneth cells are important for immunity and control the bacteria in our intestines. Since mutating ATG16L1 in mice led to Paneth cell abnormalities, we next looked at humans with the risk variant of ATG16L1. 100% of Crohn's disease patients with the ATG16L1 risk variant had similar Paneth cell abnormalities compared to 0% of the patients without the risk variant. Therefore, by comparing mice and humans, we identified a new Crohn's disease pathology and a new function for the ATG16L1 gene.
Previously, we discovered that laboratory mice are naturally infected with a mouse norovorius (MNV). Human noroviruses cause gastroenteritis (stomach flu). To determine if MNV has an effect on ATG16L1 mutant mice, we generated mice that were free of MNV infection. Surprisingly, ATG16L1 mutant mice no longer developed intestinal disease. Re-infecting the mice with MNV triggered intestinal disease. In contrast, infecting non-mutant mice with MNV did not trigger intestinal disease. Therefore, we identified a novel interaction between a viral infection and a mutation in a Crohn's disease susceptibility gene. We are excited by these results because they suggest that genetic variations associated with Crohn's disease are only dangerous when they are combined with an infectious agent. Based on our findings as well as clinical evidence, we believe that Crohn's disease can occur by more than one way. The way in which you acquire the disease may be dependent on what genetic variations you have. Further research in understanding how genetic and infectious factors interact with each other will help guide medical intervention and disease prevention strategies.