rs2187668 — HLA-DQA1 DQ2.5 tag

HLA-DQ2.5 — The Celiac Disease Gatekeeper

The HLA-DQA1¹¹ HLA-DQA1
Human leukocyte antigen genes encode cell-surface proteins that present peptides to immune cells. Variations determine which foreign and self-proteins your immune system can recognize gene encodes one chain of the HLA-DQ protein complex, which sits on the surface of antigen-presenting cells and determines what peptide fragments get shown to T cells. The rs2187668 SNP is a tag variant²² tag variant
A "tag SNP" doesn't cause disease itself but travels with disease-causing variants due to linkage disequilibrium, serving as a convenient marker that efficiently identifies the HLA-DQ2.5 haplotype—a specific combination of alleles that encodes the DQ2.5 protein isoform. This isoform has an unusually strong affinity for presenting gluten peptides to immune cells, making it the single strongest genetic risk factor³³ single strongest genetic risk factor
OR 7.04 in the initial GWAS; homozygotes have OR >10 for celiac disease.

The Mechanism

HLA-DQ2.5 consists of an alpha-5 chain (from DQA1*05) and a beta-2 chain (from DQB1*02), forming a heterodimer⁴⁴ heterodimer
A protein complex made of two different subunits on the cell surface. When gluten enters the intestine and is partially digested, certain proline-rich peptides⁵⁵ proline-rich peptides
These resist complete breakdown by digestive enzymes, making them unusually persistent escape degradation. In the intestinal lining, the enzyme tissue transglutaminase⁶⁶ tissue transglutaminase
An enzyme that normally repairs tissue damage but inadvertently modifies gluten peptides in ways that increase their immune reactivity deamidates these peptides, converting glutamine residues to glutamic acid. This modification dramatically increases their binding affinity for DQ2.5. The DQ2.5-gluten complex then activates CD4+ T cells, triggering an inflammatory cascade that damages the intestinal villi. The rs2187668 T allele tags this entire haplotype with r² = 0.97⁷⁷ r² = 0.97
Nearly perfect linkage disequilibrium, meaning the T allele almost always travels with the full DQ2.5 haplotype.

The Evidence

Genome-wide association studies⁸⁸ Genome-wide association studies
van Heel et al. tested 310,605 SNPs in 778 celiac cases and 1,422 controls identified rs2187668 as the most strongly associated variant (P < 10⁻¹⁹), with the A (also written as T on the forward strand) allele present in 53% of cases versus 14% of controls. One or two copies of DQ2.5 were present in 89% of UK celiac patients⁹⁹ 89% of UK celiac patients
Compared to 26% of population controls versus 26% of controls. Validation studies¹⁰¹⁰ Validation studies
Monsuur et al. genotyped 729 individuals confirmed that rs2187668 predicts DQ2.5 with 99.6% sensitivity and 99.4% specificity—only 7 of 1,460 chromosomes gave false results.

The gene-dose effect is substantial. Homozygous DQ2.5 carriers¹¹¹¹ Homozygous DQ2.5 carriers
Vader et al. studied T cell responses in patients with different DQ2 configurations show stronger and broader gluten-specific T cell responses than heterozygotes. In celiac patients¹²¹² celiac patients
Meta-analysis of 7 studies with 3,209 cases and 7,358 controls homozygous DQ2.5 confers approximately 5-fold higher risk than heterozygous. Mechanistic studies¹³¹³ Mechanistic studies
Megiorni et al. demonstrated preferential expression of DQ2.5 alleles revealed that DQA1*05 and DQB1*02 are expressed at much higher levels than non-predisposing alleles, explaining why even heterozygotes have high cell-surface DQ2.5 density.

Beyond celiac disease, rs2187668 associates with type 1 diabetes¹⁴¹⁴ type 1 diabetes
Howson et al. tested 24 loci in 1,384 adult-onset autoimmune diabetes cases (OR 5.36 for the T allele in type 1 diabetes), autoimmune hepatitis¹⁵¹⁵ autoimmune hepatitis
German cohort of 106 patients (24.5% T allele frequency in cases vs. 7.2% in controls, P < 0.001), and idiopathic membranous nephropathy¹⁶¹⁶ idiopathic membranous nephropathy
Meta-analysis of 11 studies with 3,209 cases (OR 3.34 for the A allele). The T allele also shows positive association¹⁷¹⁷ positive association
Type 1 Diabetes Genetics Consortium study of 6,556 cases with GAD antibodies and tissue transglutaminase antibodies in type 1 diabetes families, reflecting shared autoimmune mechanisms.

Practical Implications

The critical insight: HLA-DQ2.5 is necessary but not sufficient for celiac disease. About 25-30% of Europeans¹⁸¹⁸ 25-30% of Europeans
Population frequency estimates from multiple cohorts carry at least one copy of DQ2.5, but only 1% develop celiac disease. This means the T allele identifies genetic susceptibility, not destiny. However, the negative predictive value¹⁹¹⁹ negative predictive value
The probability that someone without the risk alleles will not develop the disease is excellent—absence of DQ2.5 (and DQ8, tagged by rs7454108) makes celiac disease extremely unlikely, useful for ruling out the diagnosis in ambiguous cases.

For dietary decisions, genetic testing alone is insufficient. Celiac disease requires serological testing²⁰²⁰ serological testing
Anti-tissue transglutaminase IgA antibodies are the first-line screen (anti-tissue transglutaminase antibodies) and, if positive, small intestine biopsy²¹²¹ small intestine biopsy
Gold standard showing villous atrophy, crypt hyperplasia, and increased intraepithelial lymphocytes showing villous atrophy. Genetic testing is most useful when serological results are equivocal, when someone is already following a gluten-free diet (antibodies disappear but genes don't), or for family members deciding whether screening is warranted.

If you're TT (homozygous DQ2.5), you have the highest genetic risk, but environmental factors—possibly including gut microbiome composition²²²² gut microbiome composition
Recent studies link specific bacterial strains to celiac risk, timing of gluten introduction in infancy, and viral infections—determine whether disease develops. Monitor for symptoms (chronic diarrhea, bloating, iron-deficiency anemia, dermatitis herpetiformis) and discuss antibody screening with your physician if symptoms arise or if you have a first-degree relative with celiac disease.

Interactions

Gene-dose effects are well documented. Compound heterozygotes²³²³ Compound heterozygotes
Individuals with DQ2.5/DQ2.2, who have two copies of DQB1*02 but only one copy of DQA1*05 with DQ2.5 on one chromosome and DQ2.2 (tagged by rs2395182 and rs7775228) on the other have intermediate risk between DQ2.5 homozygotes and simple heterozygotes, because they can form trans-heterodimers with increased DQ2.5-like function. Similarly, individuals with DQ2.2 and DQ7 (rs4639334) in trans can form DQ2.5-equivalent molecules²⁴²⁴ DQ2.5-equivalent molecules
The alpha chain from DQ7 combines with the beta chain from DQ2.2 to functionally mimic DQ2.5 cross-chromosomally, explaining celiac disease cases in people who appear DQ2.5-negative on single-SNP testing.

The combination of DQ2.5 with DQ8 (rs7454108) confers additive risk²⁵²⁵ additive risk
Each haplotype contributes independently; together they increase risk beyond either alone for both celiac disease and type 1 diabetes. In type 1 diabetes, DQ2.5/DQ8 heterozygotes represent the most common high-risk genotype²⁶²⁶ high-risk genotype
Especially in late-onset and latent autoimmune diabetes in adults, highlighting convergent autoimmune pathways.