Regulation of Host-Microbe Interactions in Autoimmunity and Antiviral Immunity by Cytosolic Nucleic Acid Sensing and Interferon Signaling

Cytosolic nucleic acid sensing and interferon (IFN) signaling are central to the host immune response to microbial pathogens. However, dysregulation of immunological pathways such as these can result in devastating autoimmune disease. In order to provide a robust immune response to pathogen without causing harm to self, the host immune system must engage in a delicate balancing act, interacting with microbes and determining whether they are commensal or pathogenic. The cGAS-STING pathway is a key regulator of host-microbe interactions by cytosolic nucleic sensing and IFN signaling. Loss of function in the cGAS-STING pathway leads to increased susceptibility to pathogenic threats, including those caused by flaviviruses such as West Nile virus. However, STING gain-of-function mutations can cause autoimmune disease in both humans and mice. Here, we investigate the contributions of cytosolic nucleic acid sensing and IFN signaling to the regulation of host-microbe interactions in antiviral immunity and autoimmune disease. Flavivirus infections represent a major global public health concern. Due to the 2015-2016 Zika virus (ZIKV) epidemic in Latin America, host-pathogen interactions at barrier sites such as the placenta and cornea are a topic of heavy research interest. Furthermore, the question of whether ZIKV is unique among flaviviruses in its ability to cross the placenta and cause congenital infection remained. Using murine models and human explants, we discovered that ZIKV-related flaviviruses, WNV and Powassan virus (POWV), have the capacity to infect both the mouse and human placenta, and cause fetal demise in mice. In contrast, alphaviruses chikungunya virus (CHIKV) and Mayaro virus (MAYV) are unable to cause fetal demise in mice, and replicate poorly in the mouse placenta and in human placental explants. Whereas IFN stimulated genes (ISGs) are up-regulated in the placenta and fetus of WNV- and POWV-infected animals, CHIKV and MAYV infection did not increase ISG levels. Whether IFN signaling at the maternal-fetus interface played a role in the pathogenesis of WNV and POWV was not determined. However, subsequent studies demonstrated that type I IFN signaling is detrimental to the mouse fetus during congenital ZIKV infection. Thus, the possibility that IFN signaling contributes to viral pathogenesis in our congenital infection model remains unknown. We additionally investigated IFN signaling at another barrier surface, the cornea. Whereas type I and type II IFN signaling during corneal infection is well-studied, less is known about type III IFN (IFN-l) signaling during corneal infection. Using ZIKV, herpes simplex virus-1 (HSV-1), and the emerging severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), our study demonstrated that IFN-l signaling can be activated in the cornea to restrict viral replication. Moreover, IFN-l induces ISG expression in the cornea in a manner dependent on the IFN-l receptor (IFNlR1), and blockade of IFNlR1 renders the cornea more susceptible to infections by ZIKV and HSV-1, but not SARS-CoV-2. Type I, II, and III IFN production can be induced by signaling through the cGAS-STING pathway. Gain-of-function mutations in STING, which result in constitutive up-regulation of IFNs, can cause STING-associated vasculopathy with onset in infancy (SAVI), a rare interferonopathy characterized by pulmonary fibrosis, skin lesions, recurrent infections, and T and NK cell cytopenia. Although STING is a sensor of host- and bacteria-derived cyclic dinucleotides (CDNs), prior studies had not investigated the role of bacteria and bacterial products in the development of STING-associated vasculopathy. We discovered that STING is hyper-responsive to bacterial CDNs. Furthermore, depletion of gut microbes with an oral antibiotic cocktail (vancomycin, neomycin, ampicillin; VNA) leads to near eradication of perivascular inflammation in the lung. However, addition of metronidazole eliminates the protective effects of VNA antibiotics. Surprisingly, germ-free SAVI mice still develop autoinflammatory lung disease, but lung disease can be rescued by recolonization with Bacteroides thetaiotaomicron or Bacteroidales-enriched stool from VNA-antibiotics treated mice. Furthermore, treatment with short-chain fatty acids, common biproducts of Bacteroidales bacteria, also prevents the development of lung disease. Although we did not determine the precise immunological mechanism of protection, we demonstrated that a metronidazole-sensitive anaerobe of the order Bacteroidales has the capacity to protect against autoinflammatory lung disease in SAVI mice. Collectively, these studies have helped to improve our understanding of the ways in which cytosolic nucleic acid sensing and IFN signaling regulate host-microbe interactions within the context of antiviral immunity and autoimmune disease. Our virus studies have provided key insights into viral pathogenesis and IFN signaling at the maternal-fetal interface and the cornea. Furthermore, our discovery that microbial composition can regulate the development of interferonopathy in SAVI mice may pave the way for more host-microbe studies and lead to novel therapeutics for monogenic autoimmune diseases such as SAVI