This file is a summary of various posts made on the CELIAC List about genetics and celiac, and more specifically about HLA typing in Celiac Disease. It has been supplemented by private E-Mail with various experts in this area of CD, and has been reviewed for technical accuracy by Drs. Ludvig Sollid (University of Oslo) and Karoly Horvath (University of Maryland), two of the CEL-PRO experts. This file is formatted as a question and answer session.
Some people have had the notion that Celiac was almost totally absent in African-American and Asian populations, however clinical observations and some studies of immigrants who have added gluten-containing foods have shown surprising rates of Celiac. Therefore it is unclear if the historical low reported rates of CD were due to lack of gluten, lack of proper diagnosis, or lower genetic susceptibility.
Additional information on CD in these populations can be found at:
(a) Brueton & al Coeliac disease in Asian children in United Kingdom: in McConnell RB (ed): The Genetics of Coeliac Disease. Lancaster, MTB Press, 1981, pp. 103-108
(b) Sher & al. High risk of coeliac disease in Punjabis. Epidemiological study in the South Asian and European Populations of Leicestershire. Digestion 1993; 54: 178-182).
(c) Boudraa et al. Epidemiology of gluten intolerance in North Africa. In: Common Food Intolerances. I.: Epidemiology of Celiac disease. Dyn. Nutr. Res., Basel, 1992. vol.2, pp 64-70.
The fact that identical twins do not always both get CD implies that there are both genetic and non-genetic factors involved. The non- genetic factors are not yet known. Certain viruses or stress triggers are suspected. The fact that there is a difference in the risk factors between identical twins (70%) and HLA-identical siblings (30-40%) implies that there are also non-HLA genetic factors involved.
Even if there are no obvious gastrointestinal symptoms, some gastroenterologists recommend that all first degree relatives of a Celiac get endomysial antibody blood tests, which are over 90% predictive of whether a biopsy will be positive. If the test is negative, it is often recommended that first degree relatives get retested 2-5 years in the future, or if symptoms associated with Celiac Disease (or unexplained symptoms) ever develop.
It appears that HLA types are not the complete, final answer on the genetics of CD. Periodically, CD patients are found who do not carry any of the known HLA risk factors which we will be discussing in this file. Thus, Celiac is a multifactorial disorder with very strong (but not exclusive) association with a few HLA factors.
In addition, there is some evidence that gluten intolerance may be a strong factor in the development or aggravation of certain developmental disorders, such as autism. These disorders typically develop during the first 2-3 years of life. When checking for the HLA markers for increased risk of Celiac, families at risk for autism might also check to see if their infant has any of the HLA markers associated with an increased risk for autism (for more information on this topic, send an E-Mail to listserv@LISTSERV.ICORS.ORG with the body:
GET CELIAC HLA-WAR1
GET CELIAC HLA-WAR2
If gluten is added after language and social interaction have become well-developed, it may dramatically reduce the probability of developing these disorders. While no data has been published to support this theory, those with an autistic child in the family can attest that maintaining a gluten-free diet is much less difficult than the problems associated with autism in general.
HLA molecules are involved in immune regulation. They bind to small fragments(peptides) of bacteria, virus etc. and present these to particular cells (T-cells) of the immune system. T-cells patrol the body and are educated to inspect whether the HLA molecules display peptides that are foreign to the body. If a T-cell recognizes something foreign, it sends out an "alarm". This will result in an immune response and the immune system will start to attempt to defeat the presumed intruder in the body. It is likely that in Celiacs the immune system reacts to gluten and that the immune response which is initiated in the small intestine on contact with gluten results in alterations of the normal gut tissue.
Why only some individuals react to gluten is not completely understood. There exist a large number of different variants of HLA molecules. Which HLA molecules an individual manufactures is determined by the gene-variants the individual has inherited from his or her parents. It is conceivable that some particular varieties of HLA molecules (i.e. DQ2 and DQ8; see later) are particularly efficient in binding and presenting gluten fragments and that individuals who carry these particular HLA molecules are more prone to develop Celiac Disease. It is important to remember that the very same HLA molecules that present gluten fragments also are very efficient in presenting fragments of virus and bacteria and are thus of great importance for the body to survive from various infections.
Most large commercial laboratories can perform HLA typing. The cost varies, depending on where it is performed, and how many HLA markers are specified to be analyzed. The costs range from $160 to $800 in the few examples the Listowners have seen. Those who donate blood for celiac genetics studies typically get the HLA typing done for free.
Not only will insurance generally NOT cover HLA typing (at least in the US), but some patients are fearful that if their insurance company discovers that they have HLA markers implying a much higher risk for developing a serious disease, that in the future, coverage may be denied. This is topic on which heated discussion is occurring in the U.S. Some patients actually use a different doctor for their HLA typing and do not submit the invoice to their insurance company for this reason.
HLA molecules exist in two classes. HLA-A, HLA-B and HLA-C molecules make up the class I molecules and HLA-DR, HLA-DQ and HLA-DP molecules make up the class II molecules. All HLA molecules consist of two units; an alpha chain and a beta chain. For each chain there is a corresponding gene. The genes coding for HLA molecules (except the gene coding for the beta chain of class I molecules) are positioned on a stretch of chromosome 6 like pearls on a string. For each gene locus (each position) there exist a large number of variants or alleles (colors of pearls). For some loci there are more than 100 different variants described! Each individual has two chromosomes (pearl strings) which are inherited from the mother and the father. The composition of alleles on the two chromosomes (combination of colored pearls on the strings at different positions) are characteristic for each individual. Typically, the genes of one chromosome (the colored pearls on one string) code for one HLA-A molecule, one HLA-B molecule, one HLA-C molecule, two HLA-DR molecules, one HLA-DQ molecule and one HLA-DP molecule. Provided that the genes on the individual's two chromosomes code for different variants (i.e. there are different color of the pearls in the corresponding positions on the two strings) an individual may express as much as 14 different HLA molecules. (The number can be larger as in some instances alpha- and beta-chains of HLA-DQ and HLA-DP molecules that are coded for by genes on opposite chromosomes may form functional molecules).
Which alleles (color of pearls) that occur together at the various loci (positions) are not completely random. Some combinations are particularly frequent. Several combinations of alleles (color of pearls) are usually maintained together as gene blocks (segments of pearls). If you have one particular allele (color of pearl) at one locus (position) you very often have another distinct allele (color of pearl) at a neighboring locus (position). This phenomena is termed linkage disequilibrium and is very important for understanding association of certain diseases with particular HLA markers.
Those Celiacs who are not DR3 positive very often are DR5/DR7 (DR5 is coded for by genes on the one chromosome and DR7 is coded for by genes on the other chromosome). Individuals who are DR3, almost always also are DQ2 (because the genes coding for DR3 and DQ2 are in linkage disequilibrium - they are in a conserved gene block). The names of the genes in this gene block that jointly code for DQ2 are DQA1*0501 (which codes for the alpha chain) and DQB1*02 (which codes for the beta chain). The DQA1*0501 gene also exists in a gene block that contains the gene coding for DR5, and the DQB1*02 gene exists in a gene block that contains the gene coding for DR7. Individuals who are DR5/DR7 may thus also possess the DQA1*0501 and DQB1*02 genes. In contrast to individuals who are DR3-DQ2 the DQ genes in DR5/DR7 individuals which code for DQ2 are located on two different chromosomes. This explains why DR5/DR7 individuals also have a predisposition to develop Celiac Disease and strongly indicates that it is DQ2 coded for by the DQA1*0501 and DQB1*02 genes that actually are involved in the development of Celiac Disease. About 95% of the Celiac population carry these particular genes.
In addition, another 4-5% of the Celiac population have genes that code for DR4 and DQ8 molecules (the genes coding for DR4 and DQ8 are maintained in block of genes). Probably it is the DQ8 molecule that mediates the disease predisposition in this subgroup of patients. Some very few Celiacs are neither DQ2 nor DQ8 implying that HLA typing is not a perfect predictor of risk for Celiac Disease.
Any disorder which is linked with DR3 is likely to be found more often amongst Celiacs than amongst the general population. By far the most well known such disorder is Type I ("juvenile") diabetes. Diabetics are almost always associated with either DR3 or DR4. Therefore, every diabetic should realize that they have a much higher risk of having, or developing, Celiac Disease than the general population. And Celiacs should realize that they may have a higher risk of developing Type I diabetes than the general population.
Other disorders which have been claimed to be associated with DR3 include: Grave's disease, Addison's disease, Sjogren's syndrome and Systemic Lupus Erythematosus. Reference: Rose et al, Manual of Clinical Laboratory Immunology, 4th Edition (1992), American Society for Microbiology, Washington DC.
Finally, those with Down's Syndrome have been reported to have both an increased risk for CD, and also an increase in gluten-sensitivity (but not Celiac Disease itself). See:
They also have a page on DH:
The purpose of this copyright is to protect your right to make free copies of this paper for your friends and colleagues, to prevent publishers from using it for commercial advantage, and to prevent ill-meaning people from altering the meaning of the document by changing or removing a few paragraphs.