Background: 
GLP-1 and GLP-2 (glucagon-like peptide-1/2) are gut hormones secreted in response to food. While GLP-1 induces postprandial insulin secretion, GLP-2 enhances intestinal nutrient absorption and is clinically used for the treatment of patients with short bowel syndrome. Recently, GLP-2 has been found to be upregulated in patients with colitis through as of yet unknown mechanisms. The aim of this study was to characterize the role of GLP-2 during systemic inflammation. 

Methods and Results: 
Therefore we measured circulating GLP-2 levels in 2 clinical cohorts under inflammatory conditions. In the first cohort (34 patients) we found GLP-2 levels to be increased over time in response to cardiac surgery as an inflammatory stimulus (from 2.9 ± 1.4 ng/mL at baseline to a maximum of 5.1 ± 2.8 ng/mL 12h post-surgery). In the second cohort 223 critically ill patients with sepsis showed a 3.9 fold increase of total GLP-2 plasma levels in comparison to healthy controls (3.0 ng/mL vs. 11.4 ng/mL; p<0.001). Moreover, GLP-2 levels correlated with markers of inflammation (IL-6, PCT, CRP) and were associated with mortality. To elucidate the underlying mechanisms we injected C57BL/6 mice with LPS or performed CLP (cecal ligation puncture). Induction of lethal sepsis by CLP or LPS injection increased GLP-2 secretion. Further experiments in IL1R^-/- and IL6^-/- mice demonstrated that LPS-induced GLP-2 secretion is mediated by IL-6. To identify the source of GLP-2 secretion under inflammatory conditions, we injected LPS into Gcg^-/- mice (preproglucagon; prescursor of GLP-2) with a tissue-specific reactivation of the Gcg gene in gut L-cells (GcgRA^ΔvilCre) or pancretic alpha cells (GcgRA^ΔPDX1-Cre). Importantly, LPS-induced GLP-2 secretion was blocked in GcgRA^ΔvilCre mice, while GcgRA^ΔPDX1-Cre mice showed a marked increase of GLP-2, indicating that inflammation-dependent GLP-2 production is derived from the pancreas and not from the gut. Additional in-vitro and ex-vivo approaches revealed that IL-6 directly activates GLP-2 secretion from murine pancreatic alpha cells and human islets. Finally, we analyzed whether inflammatory upregulation of GLP-2 has functional immunomodulatory relevance. Upregulated circulating cytokine levels in septic Gcg^-/- and Glp2r^-/- mice suggest a protective role of endogenous GLP-2 secretion. We administered GLP-2 or saline as control per central jugular vein catheter mice who underwent CLP. GLP-2 treatment improved LV-contractility (dp/dt_max) in septic cardiomyopathy, inhibited sepsis-induced hypotension and reduced mortality (p=0.018). Mechanistically GLP-2 reduced myeloid immune cell infiltration into heart and liver tissue and decreased proinflammatory cytokine levels in various organs and the blood (TNF-α, IL-6 and IL-1β). After broad GLP-2 receptor profiling we found maximum mRNA expression in gut tissues with no expression on immune cells. By further mechanistic studies we found GLP-2 to protect sepsis-induced gut barrier dysfunction. 

Conclusion:
By using a translational approach (clinical studies, experimental in vivo and in vitro studies) we identified that acute inflammation activates pancreatic GLP-2 secretion independent of food intake in mice and humans. Inflammation-induced GLP-2 activation protects against sepsis and improves survival. GLP-2 provides a previously unknown link between the pancreas and the immune system, which might open new avenues for the treatment of patients with sepsis.