Poster Presentation International Celiac Disease Symposium 2026

Gluten-like grain proteins activate coeliac disease-relevant T cell responses but show limited in vivo reactivity (#360)

Angela Juhasz 1 , Utpal Bose 1 , Melinda Y Hardy 2 , Sophia Escobar-Correas 3 , Mitchell Nye-Wood 1 , Cynthia Kanagasundaram 2 4 , Jason Tye-Din 2 4 , Michelle Colgrave 1
  1. Australian Research Council Centre of Excellence for Innovations in Peptide and Protein Science, School of Science, Edith Cowan University, Edith Cowan University, Joondalup, WA, Australia
  2. Walter and Eliza Hall Institute, Parkville, VIC, Australia
  3. Environment, Commonwealth Scientific and Industrial Research Organisation, CSIRO, Brisbane, QLD, Australia
  4. Department of Gastroenterology, Royal Melbourne Hospital, Royal Melbourne Hospital, Melbourne, VIC, Australia

Introduction

Persistent symptoms and intestinal inflammation in coeliac disease (CeD) despite a gluten-free diet (GFD) are clinically challenging. Although often attributed to low-level gluten exposure, an alternative or additional contributor may be consumption of apparently gluten-free grains that contain gluten-like proteins with pro-inflammatory potential. Using IL-2 as a sensitive marker of CeD-relevant T cell activation, we investigated whether these proteins contribute to non-responsive CeD.

 

Methods

Proteins were extracted from 16 crops consumed in the GFD including cereals, legumes, pseudocereals, and emerging crops. Gluten-like proteins and candidate immune-reactive epitopes were identified by proteomic and in silico analyses. Proteins were assessed for T cell reactivity in HLA-DQ2.5+ CeD patients by IFN-gamma ELISpot (n=10). Peptides with homology to known gluten epitopes were then tested using a CeD-relevant whole-blood IL-2 assay (n=16). Oral grain challenge was performed in 20 CeD patients with serum IL-2 measurement.

 

Results

Gluten-like protein orthologues with candidate immunoreactive peptides were identified in millet, teff, and sorghum. IFN-gamma T-cell responses to protein extracts were observed in 8/10 patients and were enhanced by deamidation but not significantly reduced by HLA-DQ blockade. Whole-blood IL-2 responses to a millet-derived peptide were observed in 9/16 patients. To assess whether this in vitro reactivity translated in vivo, oral millet flour challenges were performed: 1/20 patients showed increased serum IL-2 after ingestion, although this response was not reproduced on two repeat challenges.

 

Conclusions

Gluten-like peptide sequences in grains used in the GFD can activate cross-reactive T cells in some CeD patients, including a millet-derived peptide that induced IL-2 responses in >50% of CeD patients tested. However, controlled millet challenge did not reproducibly induce in vivo immune activation or major symptoms. These findings distinguish in vitro T cell cross-reactivity from clinically meaningful responses and suggest that millet is unlikely to contribute substantially to non-responsive CeD in most patients.