Can a Common Stomach Bug Turn Gluten into the Bad Guy?

Based on: Murine Norovirus Infection Induces Th1 Inflammatory Responses to Dietary Antigens

A little background on Celiac disease

Celiac disease (CeD) is a genetically inherited autoimmune disorder that affects about 1% of the United States population.  In people with CeD, a protein found in grains called gluten triggers the immune system to attack the person’s own small intestine.  The gluten protein can’t be digested fully, so our bodies break it down to a smaller 33 amino acid segment that normally gets expelled without causing any problems (because the normal immune system usually learns to tolerate harmless antigens like gluten).  This short piece of gluten can then be modified by an enzyme called tissue transglutaminase (tTG).  When these gluten fragments make it past the cells that make up the small intestine (called epithelial cells) in CeD patients, however, they cause an inflammatory immune response by stimulating immune cells underneath called intraepithelial lymphocytes (IELs). 

            To activate these immune cells, specialized cells called dendritic cells (DCs) have to present the gluten fragments to them.  Our DCs in our immune systems use a molecule called the major histocompatibility complex (MHC) to display different short proteins to B and T cells (for a short video about MHC presentation, follow this link!).  This presentation activates sets of helper T cells (also known as CD4+ T cells or Th cells) called Th1 and Th17 cells that secrete inflammatory molecules.  Though these molecules can directly damage the epithelial cells, they can also affect other immune cells. Killer T cells (CD8+ T cells) receive these signals from the Th1 cells and attack the epithelial cells.  Additionally, signaling from the Th cells tells B cells to make a specific type of antibody called IgA that can be directed against the TTG enzyme or the gluten fragment itself.  These antibodies can stay in the intestine and cause inflammatory responses, but some also enter the bloodstream and can be used to diagnose CeD via a simple blood test.

Immune mechanism responsible for Celiac disease (Kagnoff, 2007)

            The gut tissue is fairly sensitive, and the highly inflammatory Th1-driven response severely damages the small intestine’s ability to absorb nutrients.  As a result, CeD patients often have a multitude of nutritional deficiencies and iron-deficiency anemia.  Children with CeD often fail to follow a normal growth curve and display dental enamel defects.  The autoimmune attack also causes symptoms like diarrhea or constipation, bloating, cramping, and central nervous system issues (“brain fog”, anxiety, or depression).  If you think you or someone you know could have CeD (or you have an immediate family member with CeD), be sure to look at the symptom checklist provided by Beyond Celiac and get tested!

Celiac Genetics

            Approximately 80 to 95% of people with Celiac carry a specific sequence of MHC called HLA-DQ2 (Sollid and Thorsby, 1993), and many other patients carry the HLA-DQ8 form.  These genes are necessary for a person to develop CeD because they make it easier for the DCs to present the tTG-modified gluten to B and T cells.  The HLA-DQ2 and -DQ8 genes are both required for the development of CeD; however, approximately 40% of Western populations are estimated to carry at least one of those predisposing forms (Fasano, 2009).  Despite this high prevalence, few of these carriers actually develop CeD.  Researchers are currently interested in identifying any other genes or environmental factors that are necessary for the disease to develop in only some of the people carrying the HLA-DQ2 and/or -DQ8 genes.  Though it usually resolves all of a person’s symptoms and is therefore an effective treatment, the gluten-free diet can be difficult to maintain and have a negative emotional impact (White et al., 2016; Meyer and Rosenblum, 2017).    Since we can’t control our genetics, discovering a causal environmental factor would allow for the development of prevention methods for CeD.

Representation of the HLA allele distribution (Kagnoff, 2007) 

What does norovirus have to do with Celiac?

            One such environmental factor might be infection with some types of viruses that circulate commonly among humans.  Currently, it is unknown why people with CeD mount a Th1 response to ingested gluten instead of the normal tolerance.  The Type 1 Lang form of reovirus has been shown to influence loss of immune system tolerance to gluten by promoting the inflammatory Th1 response that is characteristic of CeD (Bouziat et al., 2017).  Though from a different family, Bouziat et al. hypothesized that another virus called norovirus might be able to cause similar effects.  Human norovirus is a common RNA virus that is responsible for most cases of the “stomach bug” around the world (Glass et al., 2009).  You may recognize the name from news stories covering outbreaks of illness on cruise ships.  The murine (mouse) norovirus (MNV) is often used as a model for the human version in animal studies, and Bouziat et al. used two different strains of murine norovirus to investigate whether they could instigate a Th1-skewed immune response to orally consumed protein.  The CW3 strain of MNV causes severe but short-duration infection, while the CR6 strain causes chronic infection. 

The Experiments

            To figure out whether MNV is capable of inducing loss of tolerance to dietary molecules, the researchers began by using a protein called ovalbumin (OVA).  The immune system of mice usually tolerates OVA when it is introduced orally, much like gluten in food typically induces tolerance in humans.  Normal mice were given one of four possible solutions orally: inert medium (as a control), OVA only, OVA and the CW3 strain of MNV, or OVA and the CR6 strain.  The mice were then injected with OVA under the skin, and the researchers evaluated whether their immune systems mounted an inflammatory response by measuring the amount of IgG2c antibodies (which are made during a Th1 response) against OVA in the blood and swelling of the ear after injection.  After receiving OVA orally, the mice should tolerate the protein and therefore have low numbers of antibodies/little swelling after the injection.  This was true for mice who received the control ("sham"), OVA alone, and OVA plus the CR6 strain.  The mice who received OVA and CR3, however, had a significant increase in IgG2c antibodies and more swelling compared to the other three groups.  These results indicate that the CW3 strain of MNV is capable of influencing the immune system to respond to dietary OVA instead of tolerating it.

(D) Higher levels of anti-OVA antibodies in mice who got CW3, (E) more ear swelling in CW-3 infected mice

            Next, Bouziat et al. wanted to determine if CW3 is able to activate a certain type of DC living in the lymph node near the gut that’s involved in the loss of tolerance to food molecules.  They followed the same procedure of oral exposure and injection described before but then counted how many of these DCs produced interleukin 12 (IL-12).  IL-12 is only produced by activated DCs and promotes the Th1 response; therefore, increased amounts of IL-12 indicate a higher likelihood of mounting a Th1 response instead of a tolerogenic response.  The CW3 strain, but not the CR6 virus, did activate more DCs (as indicated by more cells making IL-12) after injection of OVA.

            To gain a better understanding of whether a Th1 response to OVA is actually induced by CW3 or CR6, the authors transferred T cells that are specific to OVA into normal mice.  These OVA-specific naïve T cells are called OT-II and are used because they are easy to track (they express a different version of a cell surface protein than the recipient mouse’s own cells) and the scientists know what antigen they respond to.  After injection of OT-II cells, mice orally received either CW3, CR6, or no virus (sham).  These infected recipient mice were then fed an OVA-containing diet for 6 days, and the Th1 response was measured at that point (see the timeline below!).  The authors measured the Th1 differentiation by counting the percentage of OT-II cells that expressed interferon gamma (IFN-γ); IFN-γ is the main communication molecule produced by Th1 cells.  Mice that received CW3 polarized the most OT-II cells to be Th1 cells in response to dietary OVA, as indicated by higher levels of IFN-γ  production.  This result was not observed for mice who received CR6, indicating that the capability to polarize a Th1 response is specific to CW3. 

(C) Timeline of OT-II transfer experiment, (D) greater IFN-γ production by OT-II cells after CW3 infection, indicating a strong Th1 response

            The authors then wanted to determine exactly which part of the CW3 virus is important for triggering the inflammatory Th1 response.  Because the CW3 strain is able to cause inflammation but the CR6 strain is not, the viral protein that promotes the Th1 response must be a protein that the two strains don’t share.  One unshared protein is viral protein 1 (VP1), which is attached to the outer covering of both viruses but in slightly different forms.  To investigate whether the CW3 version of VP1 is responsible for its inflammatory action, the authors ran the same two experiments but swapped CR3’s own version of VP1 with the CR6 version.  In other words, all of the genes in the virus were from the CR3 strain except the VP1 gene, which was swapped for the CR6 VP1 gene (the authors called this mutant virus CW3-VP1^CR6).  When animals were fed OVA along with either a sham, normal CW3, or CW3-VP1^CR6 and then injected with OVA, CW3-VP1^CR6 failed to activate the DCs important for causing loss of tolerance (did not increase IL-12 production).  The normal CW3 did activate these DCs as it did before.  In the transfer experiment with OVA-specific OT-II cells, CW3-VP1^CR6 was unable to increase IFN-γ production by the transferred cells in response to dietary OVA.  These data suggest that without its natural VP1 protein, CW3 is unable to activate DCs and polarize a Th1 response to dietary antigen.  VP1 is therefore vital to CW3’s ability to trigger the inflammatory response.

DCs of mice who received CW3-VP1 CR6 don't produce IL-12 in response to injected OVA (right), and transferred OT-II cells from these mice don't produce IFN-γ (left) 

            Since Bouziat et al. demonstrated that the VP1 protein is necessary to induce loss of tolerance, they wanted to also investigate whether VP1 alone is sufficient to confer the ability to stimulate inflammation.  They ran the same two experiments again; however, this time, the mice either received a sham infection, CW3, CR6, or a mutant CR6 expressing the CW3 version of the VP1 protein (which they called CR6-VP1^CW3).  Essentially, this mutant is just the opposite of the one used in the previous round of experiments: all of the genes are from CR6 except for VP1, which was swapped for the CW3 version.  Mice who orally received CR6-VP1^CW3 along with OVA and were then injected with OVA had a higher number of activated DCs (produced more IL-12) than mice who were infected with regular CR6.  In the OT-II T cell transfer experiment, mice who were infected with CR6-VP1^CW3 converted more of the transferred T cells to Th1 cells as indicated by expression of the transcription factor that dictates Th1 development.  Though the mutant virus didn’t induce a Th1 response as strongly as the normal CW3 virus, it produced more of a response than the normal CR6 strain.  Taken together, these results demonstrate that the CW3 form of the VP1 gene is both necessary and sufficient to cause loss of tolerance to dietary antigen.

            Both the CW3 strain of MNV and the Type 1 Lang strain of reovirus induce an inflammatory loss of tolerance to food molecules; therefore, the authors wanted to evaluate whether the viruses turn on similar gene pathways in the host cells of the lymph node that protects the intestines.  They found that there were six gene pathways that were commonly induced in response to infection with Type 1 Lang and CW3 (but not the CR6 and CW3-VP1^CR6 strains that don’t cause loss of tolerance).  Three of these pathways are extremely relevant to CeD, including antigen processing and presentation, immune cell activation, and natural killer cell-mediated killing of healthy host cells. 

One of the genes involved in immune cell activation that was upregulated by both Type 1 Lang and CW3 is Irf1, which is a vital intracellular signaling molecule involved in the Th1 immune response.  To determine the importance of Irf1 upregulation, the authors performed the same two experiments in mice that could not express Irf1 at all (Irf1^-/- mice).  In these mice, CW3 did not activate the DCs that influence tolerance to dietary antigens as indicated by a failure of these cells to produce IL-12 in response to OVA.  In the OT-II transfer experiment, CW3 did not induce an increase in IFN-g production in Irf1^-/- mice.  These results suggest that Irf1 is important in the viruses’ triggering of the Th1 immune response after exposure to dietary antigen.

Genes upregulated in the lymph node that protects the gut in response to CW3 and Type 1 Lang infection (red squares indicate upregulation)

So what?

            Bouziat et al. provided evidence that the CW3 strain of the mouse form of norovirus is capable of pushing the immune system to mount a Th1 response against food molecules instead of tolerating them.   If norovirus infection is shown to be instrumental in causing a break of tolerance to gluten in humans, immunization against the virus may be a viable prevention strategy for those carrying the HLA-DQ2 and/or DQ8 forms of the MHC genes.  An epidemiological study looking at whether norovirus infection increases a genetically-predisposed child’s risk of actually developing CeD would be helpful to provide evidence that there may be a connection between norovirus infection and CeD in humans.  Fortunately, there are multiple other CeD treatments currently under development.  Nexvax2 has moved into Phase II clinical trials and preliminarily seems like it may cure CeD by reprogramming T cells!

References

Bouziat R et al. (2017) Reovirus infection triggers inflammatory responses to dietary antigens and development of celiac disease. Science 356:44–50 Available at: http://www.sciencemag.org/lookup/doi/10.1126/science.aah5298 [Accessed November 29, 2018].

Fasano A (2009) Surprises from Celiac Disease. Sci Am 301 Available at: https://www.jstor.org/stable/pdf/26001498.pdf [Accessed November 26, 2018].

Glass RI, Parashar UD, Estes MK (2009) Norovirus gastroenteritis. N Engl J Med 361:1776–1785 Available at: http://www.nejm.org/doi/abs/10.1056/NEJMra0804575 [Accessed November 29, 2018].

Kagnoff MF (2007) Celiac disease: pathogenesis of a model immunogenetic disease. J Clin Invest 117:41–49 Available at: http://www.jci.org/cgi/doi/10.1172/JCI30253 [Accessed November 29, 2018].

Meyer S, Rosenblum S (2017) Activities, Participation and Quality of Life Concepts in Children and Adolescents with Celiac Disease: A Scoping Review. Nutrients 9 Available at: http://www.ncbi.nlm.nih.gov/pubmed/28837103 [Accessed November 27, 2018].

Sollid LM, Thorsby E (1993) HLA susceptibility genes in celiac disease: genetic mapping and role in pathogenesis. Gastroenterology 105:910–922 Available at: http://www.ncbi.nlm.nih.gov/pubmed/8359659 [Accessed November 29, 2018].

White LE, Bannerman E, Gillett PM (2016) Coeliac disease and the gluten-free diet: a review of the burdens; factors associated with adherence and impact on health-related quality of life, with specific focus on adolescence. J Hum Nutr Diet 29:593–606 Available at: http://www.ncbi.nlm.nih.gov/pubmed/27214084 [Accessed November 27, 2018].