T. T. Chun1, D. S. Heffernan1, N. Hutchins1, W. G. Cioffi1, C. Chung1, A. Ayala1 1Brown University School Of Medicine,Surgery,Providence, RI, USA
Introduction: Sepsis remains a major clinical challenge with few effective therapeutic options. The recently described innate lymphoid cells (ILCs) play an important role in sepsis. Specifically, IL-33 mediated group 2 ILCs (ILC2s) produce Th-2 associated cytokines and have a protective effect against sepsis. IL-33 is also known to interact with invariant natural kill T (iNKT) cells to release protective cytokines. Previous studies have demonstrated that ILC2s are implicated in lung inflammation, and sepsis leads to an increased IL-33 level in lung tissue homogenates. We hypothesized that IL-33 is similarly released during sepsis into the blood, acting on the gastrointestinal tract and the liver to stimulate ILC2s as well as iNKT cells. The purpose of this study was to evaluate IL-33 as a marker of sepsis in an experimental model and human subjects.
Methods: Cecal ligation and puncture (CLP) was performed in both wildtype (WT) and invariant natural killer T cell knockout (iNKT-/-) mice. IL-33 levels were measured in liver tissue homogenates, the serum and the peritoneal fluid using enzyme-linked immunosorbent assays (ELISA). Liver non-parenchymal cells were isolated and stained with antibodies to determine expression of IL-33 receptor (IL-33R). We also obtained blood from septic patients and compared their serum IL-33 levels with those collected from otherwise healthy volunteers.
Results: CLP induced elevation of liver IL-33 in both WT and iNKT-/- mice (p<0.05). Furthermore, following CLP, there was a significantly increased percentage of IL-33R positive cells in WT (2.8 vs 20.9%;p=0.038). Sepsis did not induce an increase in IL-33R expression in NKT-/- mice. There was a trend toward increase in serum IL-33 levels in WT CLP versus sham (43.3 vs 104.6pg/ml;p=0.15), a finding not observed in iNKT-/- mice. Specifically, iNKT-/- mice compared to WT demonstrated significantly lower serum IL-33 levels following CLP (3.6 vs 104pg/ml;p=0.006). Within the peritoneal cavity, CLP induced significantly higher IL-33 levels in WT mice (0.90 vs 10.8pg/ml;p=0.039). In iNKT-/- mice, this alteration of peritoneal IL-33 levels was magnified (3.6 vs 23.4pg/ml;p=0.49). In humans, patients with sepsis had significantly higher serum IL-33 levels than the normal control group (0.17 vs 1.12pg/ml;p=0.0021).
Conclusion: Together, the observations that mouse liver, serum and peritoneal IL-33 levels increase with CLP, as does non-parenchymal cell IL-33R expression, imply that the ILC2s are involved and activated by sepsis. Moreover, this may be effected by the interaction with iNKT cells. Finally, that such changes are not restricted to mice but are also evident in humans, points at novel therapeutic/pathological targets.