03.14 Sepsis induces phenotypic changes upon both hepatic and migrating invariant Natural Killer T-cells

T. T. Chun1, E. Fallon1, W. G. Cioffi1, A. Ayala1, D. S. Heffernan1  1Brown University School Of Medicine,Division Of Surgical Research, Department Of Surgery,Providence, RI, USA

Introduction: Invariant Natural Killer T (iNKT)-cells are key regulators of the immune response to peritoneal sepsis. Following induction of sepsis, iNKT-cells become activated, leave the liver and migrate to the peritoneal cavity. Program Death Receptor-1 (PD-1), a co-inhibitory check point protein, has been shown to control the activation and migration of iNKT-cells. However, little is known about the expression of PD-1 or its ligands PD-L1 or PD-L2 upon iNKT-cells. Although a role for other co-stimulatory molecules, specifically CD25 and CD40 has been demonstrated in several chronic disease states, little is known about the expression of co-inhibitory/co-stimulatory molecules upon activated iNKT-cells following polymicrobial sepsis.

Methods: 8-12 week old male C57BL/6 mice underwent sepsis (cecal ligation and puncture (CLP)) or sham procedure. Liver and peritoneal cells were harvested for analysis by flow cytometery for early (4 hours) and later (12 hours) alterations. iNKT-cells were identified using CD1d-loaded tetramer from the NIH tetramer facility. iNKT-cells were further phenotyped using monoclonal antibodies to PD-1, PD-L1, PD-L2, CD25 and CD40.

Results:As we previously observed, 4 hours following sepsis by CLP, numbers of hepatic iNKT-cells were unchanged, whereas by 12 hours, the hepatic iNKT cell population had declined, corresponding to the emergence of iNKT cells within the peritoneal cavity. With respect to early hepatic iNKT phenotype changes, 4 hours following sepsis, there was no alteration in CD25, CD40 or PD-1/PD-L1/PD-L2 expression upon iNKT-cells. However, compared to sham, by 12 hours following sepsis, although PD-1+iNKT-cell populations remained unchanged in both the liver and peritoneal cavity, it was noted that within the liver there was a marked decline in both PD-L1+iNKT-cells (49% vs 32%;p<0.01) and PDL2+ iNKT-cells (48.1% vs 27%;p<0.001). This corresponded with an increase in both PD-L1 (41% vs 83%;p<0.01) and PD-L2 expression (49% vs 87%;p<0.01) among septic mouse peritoneal iNKTcells. With respect to other molecules, there was no change in CD40 expression upon iNKT cells (32.7% vs 31.9%;p=0.9), but by 12 hours following sepsis there was a decline in both hepatic CD25+iNKT cells (54% vs 41%;p=0.03) and peritoneal CD25+iNKT-cells (62% vs 37%;p=0.01).

Conclusion:We have previously identified a direct PD1:PDL1 ligation as a vital component to iNKT cell migration in response to sepsis. Our current data further supports this concept whereby iNKT cells which remained in the liver displayed decreased PDL1 and PDL2 ligation, potentially limiting their ability to migrate, whereas peritoneal iNKT cells which have migrated, displayed increased PDL1 and PDL2 expressions. Furthermore, the decline in CD25 expression potentially allowed for greater migration of iNKT-cells. This data further expands our understanding of the mechanism by which iNKT cells migrate into the peritoneal cavity in response to surgical sepsis.