NK killing of coronavirus-infected cells

Abstract Natural killer (NK) cells are innate lymphocytes that kill tumor and virus-infected cells, independently of peptide antigens. NK cells recognize target cells through activating and inhibitory NK receptors, which interact with specific ligands on the target cell surface. The balance of activating and inhibitory receptor signaling determines whether target cells are killed. Our lab recently found that the most conserved and universally expressed NK activating receptor, NKp46, recognizes cells undergoing endoplasmic reticulum (ER) stress by binding to externalized calreticulin (ecto-CRT), an ER chaperone that translocates to the cell surface under conditions of ER stress. Ecto-CRT binds to and activates NKp46, leading to killing of tumor, senescent, and Zika virus (ZIKV)- infected cells. We also found that ER stress during ZIKV infection downregulates the major histocompatibility complex (MHC) ligands of NK inhibitory receptors. Coronavirus (CoV) replication is tightly linked to the ER. CoV mature by budding into the ER-Golgi intermediate compartment (ERGIC), where the nucleocapsid-genomic RNA complex is enclosed within a membrane containing the membrane-bound structural viral proteins, S, E, and M. These proteins are expressed on the ER. Massive production of viral proteins along with depletion of the ERGIC membrane by budding and egress of progeny virions disrupts the ER during CoV infection, as shown using murine hepatitis virus as a model system. Given our previous work showing recognition of ZIKV-infected and ER stressed cells via NKp46/ecto-calreticulin, we wondered whether CoV-infected cells may also be recognized by NK via the same interaction. Most papers studying NK responses to SARS-CoV-2 have focused on antibody- dependent cellular cytotoxicity (ADCC) and are thus measuring a function of NK cells that depends on an adaptive immune response. Three papers investigated antibody-independent NK cell interactions with SARS- CoV-2-infected target cells in vitro, and suggest that NK cell co-culture with infected targets suppresses viral replication by measuring a reduction in viral protein or RNA. However, none of these publications explored how NK recognize infected cells. In preliminary experiments I assayed for ER stress and CRT externalization during infection by the avirulent human CoV OC43. My preliminary data confirm that OC43 infection induces upregulation of ER stress genes and show that CRT externalization occurs during infection. Based on these data, I hypothesize that CoV replication causes ER stress, externalization of CRT, and downregulation of NK inhibitory receptor ligands, causing CoV-infected cells to be recognized by NKp46 and killed by NK. I will investigate this central hypothesis by measuring expression of activating ecto-CRT and inhibitory MHC ligands on the surface of CoV-infected cells (Aim 1). I will then test the ability of primary and NK cell lines to kill CoV- infected cells and determine whether the killing of infected cells, if present, is mediated by the NKp46/ecto-CRT interaction (Aim 2). I will investigate this system using both avirulent OC43 and virulent SARS-CoV-2. The proposal will further our understanding of how the innate immune system may defend against CoV infection.