Activation of the innate immune system by endoplasmic reticulum stress and proteasome inhibition

Abstract

The impact of proteotoxic stress on innate immune activation is currently unclear. Here, I investigated differential effects of ER stress and proteasome inhibition using genetically modified cell lines and primary human cells. <br /> ER stress, both through inhibition of lysosomal degradation using TN and through perturbations of calcium homeostasis using TPG, was found to induce a type-I IFN response in human cells. While ER stress-mediated secretion of CXCL-10, an IFN-inducible protein, was observed in the monocytic leukemia cell line THP-1, it was absent in HEK293T cells that do not express cGAS. In knockout cell lines generated using the CRISPR-Cas9 system, deficient in components of cytosolic DNA and RNA sensing, I observed that CXCL-10 release downstream of ER stress was mediated by cGAS, STING, and IRF3, but not by RIG-I and MAVS, in THP-1 cells. Similar results were obtained in cGAS^-/- and STING ^-/- HT29 cells, a human colon adenocarcinoma cell line. These data indicate that ER stress mediates DNA rather than RNA recognition in the cytosol and led to hypothesize that endogenous DNA would serve as DAMP for cGAS activation in this model. Indeed, ER-stress mediated CXCL-10 secretion was absent in THP-1 cells after depletion of mtDNA (rho0 THP-1 cells), whereas sensing of exogenously added DNA was intact. Presence of cytosolic DNA in ER stressor-exposed cells could be visualized using super-resolution microscopy. Moreover, mechanistically, the type-I IFN response to ER stress was shown to be dependent on the ATF6 branch of the UPR. Altogether, these findings support a model of ER stress-induced release of DNA from mitochondria mediated by ATF6, which then serves as a ligand for cGAS-STING pathway activation and leads to a type-I IFN response in human cells. My results elucidate a previously undescribed mechanism for cGAS-dependent sensing of non-DNA innate immune stimuli in infection and in sterile inflammatory conditions. <br /> Contrastingly, no type-I IFN response could be observed in cells treated with proteasome inhibitors such as BTZ. However, exposure to these agents led to IL-1beta release in THP-1 cells and in human PBMC. BTZ-dependent IL-1beta release in PBMC and BTZ-mediated cytotoxicity occurred at similar concentrations, indicating a potential role for this phenomenon during MM treatment in vivo. The IL-1beta response was abrogated in CASP1 ^-/- THP-1 cells, and ASC specking was observed following proteasome inhibition. Furthermore, IL-1beta release was preserved, albeit significantly reduced, in NLRP3 ^-/- THP-1 cells, and treatment with the NLRP3 small molecule inhibitor MCC-950 did not significantly affect IL-1beta levels in BTZ-treated PBMC. These results elucidate a currently undescribed mechanism of NLRP3-independent inflammasome activation through a small molecule. <br /> In summary, the current study adds to the understanding of perturbations in protein turnover by defining two distinct mechanisms of innate immune activation by ER stress and inhibition of the proteasome. Results from this work may have implications for innate immune responses in infection, sterile inflammation, and cancer and warrant further investigation in vivo.