Understanding why some people develop celiac disease while others tolerate gluten indefinitely has puzzled immunologists for decades. New mechanistic evidence suggests the answer may be encoded in immune architecture long before any dietary trigger occurs — a finding with direct implications for early risk stratification and potentially for the design of tolerogenic therapies.

Published in PNAS, this research demonstrates that HLA class II risk genes associated with celiac disease — primarily HLA-DQ2 and HLA-DQ8 — do not merely present gluten-derived peptides to T cells after disease onset. Rather, they actively sculpt the naive CD4+ T cell receptor repertoire during thymic maturation, systematically enriching for TCR clonotypes predisposed to recognizing gliadin epitopes. Using repertoire sequencing of naive CD4+ T cells from genotyped donors, the investigators show that HLA risk variants skew TCR diversity in measurable, reproducible ways, effectively pre-loading the immune system toward gluten reactivity before first exposure. This shifts the conceptual model from a purely antigen-presentation story to one where genetic predisposition operates upstream, at the level of T cell selection itself.

This finding is scientifically meaningful because it resolves a long-standing ambiguity in HLA-disease association research: whether risk genes act primarily at the level of thymic selection or peripheral antigen presentation. The answer here appears to be both, but with thymic shaping playing a more foundational role than previously credited. For the broader autoimmune field, this has implications well beyond celiac disease — similar HLA-driven repertoire biases may underlie susceptibility to rheumatoid arthritis, type 1 diabetes, and multiple sclerosis. Practically, if naive repertoire composition predicts disease vulnerability, liquid biopsy-style TCR profiling could one day inform pre-symptomatic risk assessment. Key limitations include the study's likely reliance on cross-sectional human cohorts without longitudinal disease tracking, and the challenge of establishing whether skewed repertoires are sufficient — rather than merely necessary — for pathogenesis. This is a conceptually important, potentially paradigm-shifting mechanistic contribution.