The exact role of innate immune cells upon infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and their contribution to the formation of the corona virus-induced disease (COVID)-19 associated cytokine storm is not yet fully understood. We show that human in vitro differentiated myeloid dendritic cells (mDC) as well as M1 and M2 macrophages are susceptible to infection with SARS-CoV-2 but are not productively infected. Furthermore, infected mDC, M1-, and M2 macrophages show only slight changes in their activation status. Surprisingly, none of the infected innate immune cells produced the pro-inflammatory cytokines interleukin (IL)−6, tumor necrosis factor (TNF)-α, or interferon (IFN)−α. Moreover, even in co-infection experiments using different stimuli, as well as non-influenza (non-flu) or influenza A (flu) viruses, only very minor IL-6 production was induced. In summary, we conclude that mDC and macrophages are unlikely the source of the first wave of cytokines upon infection with SARS-CoV-2.

Therapeutic monoclonal antibodies (mAbs) such as the superagonistic, CD28-specific antibody TGN1412, or OKT3, an anti-CD3 mAb, can cause severe adverse events including cytokine release syndrome. A predictive model for mAb-mediated adverse effects, for which no previous knowledge on severe adverse events to be expected or on molecular mechanisms underlying is prerequisite, is not available yet. We used a humanized mouse model of human peripheral blood mononuclear cell-reconstituted NOD-RAG1-/-Aβ-/-HLADQ(tg+ or tg-)IL-2Rγc-/- mice to evaluate its predictive value for preclinical testing of mAbs. 2–6 hours after TGN1412 treatment, mice showed a loss of human CD45+ cells from the peripheral blood and loss of only human T cells after OKT3 injection, reminiscent of effects observed in mAb-treated humans. Moreover, upon OKT3 injection we detected selective CD3 downmodulation on T cells, a typical effect of OKT3. Importantly, we detected release of human cytokines in humanized mice upon both OKT3 and TGN1412 application. Finally, humanized mice showed severe signs of illness, a rapid drop of body temperature, and succumbed to antibody application 2–6 hours after administration. Hence, the humanized mouse model used here reproduces several effects and adverse events induced in humans upon application of the therapeutic mAbs OKT3 and TGN1412.

No abstract

Upon treatment with polyinosinic:polycytidylic acid [poly(I:C)], an artificial double-stranded RNA, type I interferon receptor-deficient (IFNAR −/− ) mice develop severe liver injury seen by enhanced alanine aminotransferase (ALT) activity in the serum that is not observed in their wildtype (WT) counterparts. Recently, we showed that liver injury is mediated by an imbalanced expression of interleukin (IL)-1β and its receptor antagonist (IL1-RA) in the absence of type I IFN. Here we show that despite comparable expression levels of IL-1β in livers and spleens, spleens of poly(I:C)-treated IFNAR −/− mice show no signs of injury. In vitro analyses of hepatocytes and splenocytes revealed that poly(I:C) had no direct toxic effect on hepatocytes. Furthermore, expression levels of cytokines involved in other models for liver damage or protection such as interferon (IFN)-γ, transforming growth factor (TGF)-β, IL-6, IL-10, IL-17, and IL-22 were comparable for both organs in WT and IFNAR −/− mice upon treatment. Moreover, flow cytometric analyses showed that the composition of different immune cells in livers and spleens were not altered upon injection of poly(I:C). Finally, we demonstrated that the receptor binding IL-1β, IL1R1, is specifically expressed in livers but not spleens of WT and IFNAR −/− mice. Accordingly, mice double-deficient for IFNAR and IL1R1 developed no liver injury upon poly(I:C) treatment and showed ALT activities comparable to those of WT mice. Collectively, liver injury is mediated by the organ-specific expression of IL1R1 in the liver.

Type I interferons (IFNs) are a first line of defence against viral infections. Upon infection, a first small wave of early type I IFN, mainly IFN-β and particularly IFN-α4, are induced and bind to the type I IFN receptor (IFNAR) to amplify the IFN response. It was shown for several viruses that robust type I IFN responses require this positive feedback loop via the IFNAR. Recently, we showed that infection of IFNAR knockout mice with the orthomyxovirus Thogoto virus lacking the ML open reading frame (THOV(ML-)) results in the expression of unexpected high amounts of type I IFN. To investigate if IFNAR-independent IFN responses are unique for THOV(ML-), we performed infection experiments with several negative-strand RNA viruses using different routes and dosages for infection. A variety of these viruses induced type I IFN responses IFNAR-independently when using the intraperitoneal (i.p.) route for infection. In vitro studies demonstrated that myeloid dendritic cells (mDC) are capable of producing IFNAR-independent IFN-α responses that are dependent on the expression of the adaptor protein mitochondrial antiviral-signalling protein (MAVS) whereas pDC where entirely depending on the IFNAR feedback loop in vitro. Thus, depending on dose and route of infection, the IFNAR feedback loop is not strictly necessary for robust type I IFN expression and an IFNAR-independent type I IFN production might be the rule rather than the exception for infections with numerous negative-strand RNA viruses.