Termination of STING responses is mediated via ESCRT-dependent degradation.

cGAS-STING signalling is induced by detection of foreign or mislocalised host double-stranded (ds)DNA within the cytosol. STING acts as the major signalling hub, where it controls production of type I interferons and inflammatory cytokines. Basally, STING resides on the ER membrane. Following activation STING traffics to the Golgi to initiate downstream signalling and subsequently to endolysosomal compartments for degradation and termination of signalling. While STING is known to be degraded within lysosomes, the mechanisms controlling its delivery remain poorly defined. Here we utilised a proteomics-based approach to assess phosphorylation changes in primary murine macrophages following STING activation. This identified numerous phosphorylation events in proteins involved in intracellular and vesicular transport. We utilised high-temporal microscopy to track STING vesicular transport in live macrophages. We subsequently identified that the endosomal complexes required for transport (ESCRT) pathway detects ubiquitinated STING on vesicles, which facilitates the degradation of STING in murine macrophages. Disruption of ESCRT functionality greatly enhanced STING signalling and cytokine production, thus characterising a mechanism controlling effective termination of STING signalling.

Discordance in STING-Induced Activation and Cell Death Between Mouse and Human Dendritic Cell Populations.

Stimulator of Interferon Genes (STING) is a cytosolic sensor of cyclic dinucleotides (CDNs). The activation of dendritic cells (DC) via the STING pathway, and their subsequent production of type I interferon (IFN) is considered central to eradicating tumours in mouse models. However, this contribution of STING in preclinical murine studies has not translated into positive outcomes of STING agonists in phase I & II clinical trials. We therefore questioned whether a difference in human DC responses could be critical to the lack of STING agonist efficacy in human settings. This study sought to directly compare mouse and human plasmacytoid DCs and conventional DC subset responses upon STING activation. We found all mouse and human DC subsets were potently activated by STING stimulation. As expected, Type I IFNs were produced by both mouse and human plasmacytoid DCs. However, mouse and human plasmacytoid and conventional DCs all produced type III IFNs (i.e., IFN-λs) in response to STING activation. Of particular interest, all human DCs produced large amounts of IFN-λ1, not expressed in the mouse genome. Furthermore, we also found differential cell death responses upon STING activation, observing rapid ablation of mouse, but not human, plasmacytoid DCs. STING-induced cell death in murine plasmacytoid DCs occurred in a cell-intrinsic manner and involved intrinsic apoptosis. These data highlight discordance between STING IFN and cell death responses in mouse and human DCs and caution against extrapolating STING-mediated events in mouse models to equivalent human outcomes.