STING and TRIF Contribute to Mouse Sepsis, Depending on Severity of the Disease Model

ABSTRACTIFN regulatory factor (IRF)3 plays a detrimental role in the cecal ligation and puncture (CLP) mouse model of sepsis. However, it is unclear which pathway activates IRF3 in this context. In this report we investigate two pathways that activate IRF3the Stimulator of Interferon Genes (STING) pathway (which senses cytosolic DNA) and the TIR-domain-containing adapter-inducing interferon-β (TRIF) pathway (which sense dsRNA and LPS via Toll-like receptor (TLR) 3 and 4). Initially, we examine the impact of these pathways using a severe CLP model (∼90% mortality). Both STING-KO and TRIF-KO mice are protected from severe sepsis, exhibiting reduced mortality, disease score, hypothermia and inflammatory cytokines relative to WT counterparts. STING/TRIF-DKO mice exhibit a similar phenotype to each of the single KO strains, suggesting that these pathways have an interrelated function. Subsequently, we examine the impact of these pathways using a moderate CLP model incorporating clinical treatments (Lactated Ringerʼs Solution and antibiotics, ∼36% mortality). In this case, STING-KO mice show a similar phenotype to WT counterparts, while TRIF-KO mice show improved disease score and hypothermia. During sepsis, innate immune receptors recognize bacterial ligands and host-derived danger signals, including cell-free DNA released into the circulation. We show that IRF3 is activated in cultured macrophages treated with bacteria derived from the mouse cecum, dependent on TRIF, and in macrophages treated with mouse genomic DNA/Lipofectamine 2000, dependent on STING. Together, our data demonstrate that both the STING and TRIF pathways can promote sepsis pathogenesis; however their contribution depends on the severity of the disease model. We show that bacteria are abundant in the peritoneum following both severe and moderate CLP, while cell-free DNA is more highly elevated in the serum following severe CLP compared to sham and moderate CLP. Hence, the presence of bacteria and cell-free DNA may explain the variable phenotypes in our severe CLP model (dependent on TRIF and STING) vs. our moderate CLP model (dependent on TRIF only).