The Case of the Sugary Sophomore
Dominick got the chicken pox when he was 19 years old and a sophomore at UCLA. He missed a week of classes, but once the rash stopped itching he felt fine. However, he soon noticed that he was often thirsty and that he had to get up several times a night to urinate. When he also began to lose weight, he went to his family physician for a checkup.
Type 1 or Insulin-dependendent diabetes mellitus (IDDM) is the result of destruction of the insulin-secreting beta cells located in the Islets of Langerhans of the pancreas. Insulin is a hormone required for the uptake of blood glucose. People with diabetes have very high levels of blood glucose while having cells starved for glucose. The body attempts to deal with the high blood glucose by excreting glucose in the urine (glucosuria), accounting for the thirst and frequent urination. Diabetics also have high glucagon that signals the liver to release stored glucose, exacerbating the high blood glucose. In the absence of glucose entry into cells, ketones are formed from amino acid metabolism and used for energy. Over time, high blood glucose damages blood vessels, coronary artery disease, loss of sensory nerve signals, and damaged kidneys; IDDM can result in blindness, amputation (usually of toes or lower limbs), and kidney failure. Although the symptoms of IDDM and Type II diabetes are similar, Type II diabetes is a metabolic disease associated with being overweight.
In the 1980's, IDDM was identified as an autoimmune disease. The trigger is thought to be a virus infection; Coxsackie virus is the most common virus linked to development of IDDM. Coxsackie Virus and human Cytomegalovirus, which cause common, usually minor, infections in most children, contain proteins that share amino acid sequences with GAD. GAD, glutamic acid decarboxylase, is expressed in high amounts by beta cells in the pancreas. Newly diagnosed diabetics who were given cyclosporin A, an anti-rejection medication, no longer showed any signs of diabetes. However, when the cyclosporin was not taken, the diabetes quickly returned. At some point when all islet cells had been killed, cyclosporin will no longer prevent symptoms of diabetes. Since cyclosporin has long term side effects (damage to the kidney and liver), it is not used as a treatment for diabetes. Diabetics take humulin, a human insulin made by genetically engineered E. coli.
Figure 1. Diabetes is preceded by infiltration of T cells into the pancreas. Red = insulin staining; green = CD3 staining.
Diabetes has been studied extensively in a strain of mouse called NOD (non-obese diabetic). NOD mice spontaneously develop IDDM. It has been shown in NOD mice that even before the mice begin excreting glucose in their urine, their pancreases are infiltrated by T cells. Figure 1 is a fluorescence micrograph of an islet, stained red for insulin and green for CD3. It is believed that CD8 T cells kill the islet cells. In addition to CTL, diabetics have antibodies specific for insulin, GAD, and other beta cell proteins.
As with many autoimmune diseases, IDDM is associated with certain HLA alleles, especially HLA DR3 and DR4.A specific change in amino acid 57 in the beta chain of DQ is also associated with IDDM. As shown in Figure 2, a salt bridge normally formed between the beta and alpha chains of DQ at one end of the peptide-binding site cannot be formed if the asp57 is replaced by ala57.
Figure 2 . Position 57 of HLA-DQbeta chain related to IDDM. In non-diabetic DQbeta alleles, position 57 is an aspartic acid that can form an salt bridge with arginine in the alpha chain, stabilizing the peptide-binding site. In a diabetes-associated DQbeta chain, the aspartic acid is replaced with alanine, which cannot form the salt bridge.
Questions for Class Discussion
1. How could a virus infection trigger an immune response against pancreatic islet cells?
2. Why are there autoantibodies against so many different beta cell proteins? How could antibodies contribute to beta cell death?
3. It has been assumed that most of the damage to islet cells is done by CTL; however, IDDM is linked to HLA-DR and HLA-DQ alleles. How could you explain this finding?
4. If one identical twin develops IDDM, there is a 50% chance that the other twin will develop the disease. If one sibling develops IDDM, the chances that an HLA-identical sibling who is not an identical twin will develop IDDM is significantly less than 50%. Offer an explanation for these findings.
5. "Peripheral" self antigens are produced and presented in the thymus and other lymphoid organs. Insulin message and protein can be found in the thymus, lymph nodes, and spleen. One gene responsible for the expression of self antigens in the thymus is the AIRE gene. Mice with mutations in AIRE are born with autoimmune polyendocrine syndrome (APS-1) and have Addison's disease, hypoparathyroidism, mucocutaneous candidiasis, hepatitis, and Type I diabetes. How does antigen expression in the thymus influence the development of autoimmunity?
Coxsackie Virus and Human Cytomegalovirus might trigger IDDM by antigenic mimicry, sharing epitopes with GAD. However, other viruses suspected of triggering IDDM do not have antigen mimicry with beta cell molecules. In order to understand the possible role of virus infections in triggering IDDM, Karl Lang and coworkers did a series of experiments summarized in Figure 3.
Figure 3. Role of virus infection in IDDM.
6. Lang used transgenic mice that had an LCMV (mouse virus) nucleoprotein (NP) gene expressed under the control of the insulin promoter. Only the pancreatic beta cells of the transgenic mice express NP. These transgenic mice did not spontaneously develop IDDM. However, if the mice are infected by LCMV and LCMV-specific CD8 T cells are activated, the activated T cells enter the islets and kill the beta cells.
Why is LCMV NP expressed only in pancreatic beta cells? Why didn't the transgenic mice get IDDM if they weren't infected with LCMV?
Figure 4. Expansion of CD8+ T cells specific for beta islet cell antigen alone does not necessarily induce diabetes.
7. In a similar experiment, Lang used transgenic (RIP-GP) mice with LCMV glycoprotein (GP) expressed under the control of rat insulin promoter (RIP). Figure 4a above shows the changes in blood glucose (black squares) and frequency of CD8+ T cells specific for LCMV-GP peptide 33-41 (tet-gp33) at varying times following infection of the transgenic mice with LCMV.
8. Figure 4b shows the results if the RIP-GP mice were infected with a Vaccinia Virus (VV) carrying the LCMV GP gene. How do these results compare with those in Fig 4a?
9. Figure 4c shows the results if the RIP-GP mice were injected with gp33 (KAVYNFATM) + the adjuvant CpG to deliver optimal co-stimulation. Summarize the results of this experiment. Does the presence of LCMV gp33-specific CD8+ T cells correlate with the presence of clinical disease (diabetes)? How could there be antigen-specific Tc cells present without damage to the beta cells?
10. RIP-GP mice that had been previously immunized (primed) with gp33 did not become diabetic when infected with LCMV?
Figure 5. Prevention of LCMV-induced IDDM by lack of myD88 or IFN Type 1 receptor.
11. Bone marrow cells deficient (Myd88-/-) or expressing (Myd88+/+) the TLR adapter molecule Myd88 (necessary for signaling via TLR receptors) were adoptively transferred into lethally irradiated RIP-GP or C57BL/6 mice. The mice were then infected with LCMV.
a. Why were the mice lethally irradiated before transfer of bone marrow cells? Which cells have TLR?
b. Figure 5a shows levels of MHCI expression on the chimeric mice, and Figure 5b shows blood glucose levels in the chimeric RIP-GP mice. Is signaling through TLR necessary for increased expression of MHCI? for development of IDDM?
c. Figure 5e shows the percent of CD8 T cells producing IFNγ following re-stimulation of the chimeric mice with gp33. Does the presence of TLR signaling affect CD8+ T cell activation?
d. Figure 5f and 5g show the results of similar experiments in mice expressing (Ifnar+/+) or not expressing (Ifnar-/-) receptor for IFNα. How does this experiment expand our understanding of the role of virus infection in triggering IDDM?
Additional sources
Eisenbarth, S. C. and D. Homann. Primer: Immunology and Autoimmunity. in Type I diabetes: Molecular, Cellular, and Clinical Immunology. http://www.uchsc.edu/misc/diabetes/oxch1.html
Lang, K. S. et al. Toll-like receptor engagement converts T-cell autoreactivity into overt autoimmune disease. Nature Medicine 11: 38 - 145 (2005)