Celiac disease patient sera or total IgG was injected into athymic mice. A significant delay in weight gain and mild diarrhea was observed. Mice evinced significantly decreased villus height crypt depth ratios. Celiac disease-specific autoantibody deposits were present in several tissues. The condition in mice resembles early stage celiac disease in humans.

A characteristic feature of celiac disease is the presence of circulating autoantibodies targeted against transglutaminase 2 (TG2), reputed to have a function in angiogenesis. In this study we investigated whether TG2-specific autoantibodies derived from celiac patients inhibit angiogenesis in both ex vivo and in vivo models and sought to clarify the mechanism behind this phenomenon. We used the ex vivo murine aorta-ring and the in vivo mouse matrigel-plug assays to address aforementioned issues. We found angiogenesis to be impaired as a result of celiac disease antibody supplementation in both systems. Our results also showed the dynamics of endothelial cells was affected in the presence of celiac antibodies. In the in vivo angiogenesis assays, the vessels formed were able to transport blood despite impairment of functionality after treatment with celiac autoantibodies, as revealed by positron emission tomography. We conclude that celiac autoantibodies inhibit angiogenesis ex vivo and in vivo and impair vascular functionality. Our data suggest that the anti-angiogenic mechanism of the celiac disease-specific autoantibodies involves extracellular TG2 and inhibited endothelial cell mobility.

Phase II clinical trials are already ongoing to test the efficacy of novel alternative treatments for celiac disease. Before any of the candidates can enter Phase III trials, however, researchers must develop novel reliable non-invasive surrogate markers for intestinal injury and disease activity which also accurately reflect patient-related outcomes. Any novel medication for celiac disease should be as effective and safe as the gluten-free diet, and this constitutes a challenge for drug development. It is thus likely that the first medication entering the market will be supplementary to rather than substitute the gluten-free diet.

Coeliac disease is hallmarked by an abnormal immune reaction against ingested wheat-, rye-and barleyderived gluten and the presence of transglutaminase 2 (TG2)-targeted autoantibodies. The small-bowel mucosal damage characteristic of the disorder develops gradually from normal villus morphology to inflammation and finally to villus atrophy with crypt hyperplasia. Patients with early-stage coeliac disease have TG2-autoantibodies present in serum and small-intestinal mucosa and they may already suffer from abdominal symptoms before the development of villus atrophy. Previously we have shown that intraperitoneal injections of coeliac patient-derived sera or purified immunoglobulin fraction into mice induce a condition mimicking early-stage coeliac disease. In the current study, we sought to establish whether recombinantly produced patient-derived TG2-targeted autoantibodies are by themselves sufficient for the development of such an experimentally induced condition in immune-compromised mice. Interestingly, mice injected with coeliac patient TG2-antibodies had altered small-intestinal mucosal morphology, increased lamina propria cellular infiltration and disease-specific autoantibodies deposited in the small bowel, but did not evince clinical features of the disease. Thus, coeliac patient-derived TG2-specific autoantibodies seem to be sufficient for the induction of subtle small-bowel mucosal alterations in mice, but the development of clinical features probably requires additional factors such as other antibody populations relevant in coeliac disease.