ICAM-1 MEDIATES PLASMACYTOID DENDRITIC CELL SENSING OF SARS-CoV-2-INFECTED CELLS

ABSTRACT

Background: Plasmacytoid dendritic cells (pDCs) are the major producer of type I IFNs (IFN-I), the critically important antiviral cytokines against SARS-CoV- 2. Although pDCs can sense cell-free SARS-CoV-2 virions, it is unknown whether they can detect infected cells to produce IFN-I. Since cell-to-cell transmission accounts for 90% of SARS-CoV-2 infections (Zeng et al., 2022), we examined the relevance of pDC sensing of infected cells in SARS-CoV-2 infection and whether the virus exploits this pathway to evade IFN-I responses. Method(s) LSPQ1, the first SARS-CoV-2 clinical isolate received from the Public Health Laboratory of Quebec, was used as a prototype virus. SARS-CoV-2 variants of concerns (VOCs) were also used. PBMCs or enriched pDCs were cocultured with mock-infected or SARS-CoV-2-infected HeLa-hACE2 or Calu-3. Either PBMCs, enriched pDCs, or HeLa-hACE2 were pretreated with anti-human ICAM-1 antibody or isotype control. The conjugate formation was determined by flow cytometry. Polarized Caco cells were used to validate critical data. Result(s) Upon sensing infected cells, PBMCs release 6-fold more IFN-I than they do when exposed to cell-free virions. Antibody-mediated depletion of pDCs from PBMCs abolishes IFN-I secretion. Direct contact of pDCs with infected cells is required for sensing since the use of a transwell membrane reduces IFN-I release by 85%. Infected cells form conjugates with pDCs more frequently (3.2-fold higher) than uninfected cells. Blocking ICAM-1 on infected cells or pDCs impacts conjugate formation and significantly suppresses IFN-I production by 55-80%, suggesting bidirectional interaction. Moreover, human lung cells infected with VOCs are sensed to a different extent with the alpha variant being the least efficiently sensed by pDCs compared to the delta or omicron strains. Even though SARS-CoV-2 is primarily released from the apical domain of polarized infected Caco cells, sensing of infected cells does occur upon direct contact of pDCs with the basolateral domain, highlighting how pDCs antiviral responses might be triggered in respiratory tissues. Conclusion(s) pDC sensing of infected cells accounts for the vast majority of IFN-I released during SARS-CoV-2 infection. ICAM-1 promotes physical contact between pDCs and infected cells, thus leading to efficient sensing. Differential pDC sensing of SARS-CoV-2 VOC-infected cells suggests that some VOCs might manipulate the interactions of pDCs with infected cells to limit IFN-I responses.