Objective: Intestinal mucosal injury in celiac disease(CeD) is primarily mediated by inflammatory cells driven by gluten-specific T cells, yet the intracellular mechanisms underlying persistent villous atrophy in gluten-free diet(GFD) non-responsive CeD remain poorly understood. This study aims to establish a patient-derived enteroid model from treatment-naïve CeD, GFD-responsive CeD, and GFD-non-responsive CeD, characterizing epithelial dysfunction across these groups.
Methods: Intestinal organoids were derived from crypt basal cells of four well characterized patient groups: treatment-naïve CeD (Marsh 2-3b, high IgA anti-tTG titers,n=5), GFD-responsive CeD (Marsh 0-1, normal IgA anti-tTG titers,n=5), and GFD-non-responsive CeD (Marsh 2-3a,n=5) & controls(n=5). Enteroid morphology (area expansion, protrusions, counts), viability (live/dead assay), and expression of intestinal stem cell (ISC) niche markers (Na⁺/K⁺ ATPase, β-catenin, MUC2, Villin, Ki-67, γ-H2AX, Lgr5, β-defensin, Bmi1) were assessed via bright-field microscopy, confocal imaging, and quantified using ImageJ.
Results: Treatment-naïve CeD enteroids exhibited significantly impaired growth compared to controls, with maximal area deficits emerging during differentiation (Day 10: 29,484±2,471 vs. 36,453±6,208 µm²,p=0.048). GFD-responsive enteroids showed superior expansion over non-responsive counterparts by Day 6 (6,570±792 vs. 4,468±1,212 µm²,p=0.016), persisting through Day 10 (34,265±4,905 vs. 24,219±6,941 µm²,p=0.03). Passaging eliminated size differences between responsive and non-responsive groups by passage 3. Immunofluorescence revealed reduced villin expression in treatment-naïve CeD versus controls (19.45±6.90 vs. 34.45±7.35%,p=0.0104) and elevated γ-H2AX in non-responsive versus responsive CeD (3.02±1.45 vs. 0.64±0.24%,p=0.0068). Lgr5 was upregulated in treatment-naïve CeD versus controls (6.57±4.37 vs. 1.47±1.99%, p=0.0448), while Bmi1 was elevated in non-responsive versus responsive CeD (median0.23 vs. 0.00,p=0.0079). No significant differences were observed for Na⁺/K⁺ ATPase, β-catenin, Ki-67, or MUC2 .
Conclusion: Patient-derived enteroid models recapitulate key disease phenotypes, revealing impaired epithelial cell loss, epithelial differentiation, reduced organoid formation, and defective maturation in CeD, particularly accentuated in those with GFD-non-response. Enteroids could serve as robust ex-vivo platform for studying CeD pathobiology and mechanisms of treatment non-response.