A Mouse Model of Celiac Disease

**Background** Celiac disease (CeD) is an autoimmune-like intestinal disorder triggered by dietary gluten in genetically susceptible individuals carrying HLA-DQ2 or HLA-DQ8. Key histological features include intraepithelial lymphocytosis, crypt hyperplasia, and villous atrophy leading to malabsorption. Previous mouse models expressing HLA-DQ2 or DQ8 developed gluten-specific T-cell responses but failed to reproduce villous atrophy. The DQ8-Dd-villin-IL-15tg mouse model was engineered based on human observations: IL-15 is upregulated in the epithelium and lamina propria in >2/3 of CeD patients, and villous atrophy requires both epithelial stress and anti-gluten adaptive immunity. This model combines HLA-DQ8 expression with IL-15 overexpression in both epithelial (villin promoter) and hematopoietic (Dd promoter) compartments.

**Methods** Mice were maintained on a gluten-free diet and then exposed to gluten via oral gavage of gliadin balls (20 mg crude gliadin) every other day for 30 days. Control mice received PBS. Small intestine (distal ileum) was collected 0.5 cm from cecum, fixed in formalin, paraffin-embedded, sectioned (5 μm), and stained with H&E. Villous height and crypt depth were measured on five well-oriented villi using image analysis software; villous height/crypt depth ratio was calculated. BrdU (1 mg) was injected intraperitoneally 24 hr before euthanasia to assess epithelial cell proliferation. Intraepithelial lymphocytes (IELs) were counted per 100 epithelial cells. For flow cytometry, IELs were isolated by EDTA dissociation and Percoll gradient centrifugation. Degranulation was assessed by CD107a/b surface expression after 4 hr stimulation with PMA (50 ng/ml) and ionomycin (1000 ng/ml) in the presence of GolgiPlug and GolgiStop. Cells were stained with antibodies against CD45, TCRβ, CD4, CD8α, CD8β, and intracellular IFN-γ. TG2 activation was visualized by injecting 5-(biotinamido)-pentylamine (5-BP, 100 mg/kg) 6 and 3 hr before euthanasia, followed by immunofluorescence co-staining for 5-BP (streptavidin-Alexa594) and TG2 (rabbit anti-TG2 + anti-rabbit-Alexa488). Anti-gliadin antibodies were measured by ELISA using pepsin-trypsin digested gliadin (PT-gliadin) or chymotrypsin-digested gliadin (CT-gliadin) as coating antigens.

**Key Results** - Villous atrophy developed in approximately 70% of gluten-fed DQ8-Dd-villin-IL-15tg mice after 4 weeks of gluten exposure, with a decrease in villous height/crypt depth ratio. Crypt hyperplasia was not consistently observed. - BrdU incorporation was seen at the tip of villi in mice with villous atrophy, indicating increased epithelial cell renewal. - IEL density increased from 15-25 per 100 epithelial cells in sham-fed mice to an average of 40 per 100 epithelial cells in gluten-fed mice. - CD8+ IELs from gluten-fed mice showed increased degranulation (CD107a/b expression) and IFN-γ production upon stimulation. - TG2 was activated in the small intestine upon gluten feeding, as shown by 5-BP incorporation. - Anti-deamidated gliadin antibodies (IgG and IgA) were detected by ELISA using CT-gliadin; anti-TG2 antibodies were inconsistently detected. - Villous atrophy was dependent on CD4+ and CD8+ T cells, HLA-DQ8, and TG2. IL-15 overexpression in both epithelial and hematopoietic compartments was required; mice with IL-15 only in epithelium (DQ8-villin-IL-15tg) had expanded IELs but no tissue destruction, while mice with IL-15 only in hematopoietic cells (DQ8-Dd-IL-15tg) developed TH1 anti-gluten responses but no villous atrophy. - TG2 inhibition prevented villous atrophy in this model.

**Clinical Implications** This mouse model recapitulates key features of active CeD, including HLA- and gluten-dependent villous atrophy, intraepithelial lymphocytosis, cytotoxic IEL activation, TG2 activation, and anti-deamidated gliadin antibodies. It demonstrates that IL-15 must be upregulated in both epithelial and hematopoietic compartments to license IELs for tissue destruction. The model provides a preclinical platform for testing novel therapies, such as TG2 inhibitors, which have shown promise in clinical trials. The requirement for 10-12 mice per group due to 70% penetrance of villous atrophy should be considered in study design.