SARS-CoV-2 infection impairs NK cell functions via activation of the LLT1-CD161 axis

Abstract

Introduction: Natural killer (NK) cells plays a pivotal role in the control of viral infections, and their function depend on the balance between their activating and inhibitory receptors. The immune dysregulation observed in COVID-19 patients was previously associated with downregulation of NK cell numbers and function, yet the mechanism of inhibition of NK cell functions and the interplay between infected cells and NK cells remain largely unknown.

Methods: In this study we show that SARS-CoV-2 infection of airway epithelial cells can directly influence NK cell phenotype and functions in the infection microenvironment. NK cells were co-cultured with SARS-CoV-2 infected epithelial cells, in a direct contact with A549^ACE2/TMPRSS2 cell line or in a microenvironment of the infection in a 3D ex vivo human airway epithelium (HAE) model and NK cell surface expression of a set of most important receptors (CD16, NKG2D, NKp46, DNAM-1, NKG2C, CD161, NKG2A, TIM-3, TIGIT, and PD-1) was analyzed.

Results: We observed a selective, in both utilized experimental models, significant downregulation the proportion of CD161 (NKR-P1A or KLRB1) expressing NK cells, and its expression level, which was followed by a significant impairment of NK cells cytotoxicity level against K562 cells. What is more, we confirmed that SARS-CoV-2 infection upregulates the expression of the ligand for CD161 receptor, lectin-like transcript 1 (LLT1, CLEC2D or OCIL), on infected epithelial cells. LLT1 protein can be also detected not only in supernatants of SARS-CoV-2 infected A549^ACE2/TMPRSS2 cells and HAE basolateral medium, but also in serum of COVID-19 patients. Finally, we proved that soluble LLT1 protein treatment of NK cells significantly reduces i) the proportion of CD161+ NK cells, ii) the ability of NK cells to control SARS-CoV-2 infection in A549^ACE2/TMPRSS2 cells and iii) the production of granzyme B by NK cells and their cytotoxicity capacity, yet not degranulation level.

Conclusion: We propose a novel mechanism of SARS-CoV-2 inhibition of NK cell functions via activation of the LLT1-CD161 axis.

Keywords: CD161-LLT1 axis; NK cell impairment; NK cells; SARS-CoV-2; antiviral response.

Figure 1

Scheme of the experimental setup. NK cells were added to 72h prior SARS-CoV-2 infected A549^ACE2/TMPRSS2 cells, co-cultured for 24h and FACS analysis of expression of NK cell receptors was performed (A) NK cells were added to 48h prior SARS-CoV-2 infected HAE, co-cultured for 72h and FACS analysis of expression of NK cell receptors and cytotoxic assay against K562 cells were performed (B).

Figure 2

SARS-CoV-2 infection alters the proportion of NKG2D⁺ and CD161⁺ NK cells due to direct contact of NK cells with virus-infected A549^ACE2/TMPRSS2 cells. Flow cytometry analysis of the proportion of NK cells expressing surface receptors CD16, NKp46, NKG2D, DNAM-1, NKG2C, CD161, NKG2A, TIM-3, TIGIT, and PD-1 in NK cells co-cultured with SARS-CoV-2 and mock-infected A549^ACE2/TMPRSS2 cells. Data were obtained from five independent experiments, using NK cells isolated from five different healthy blood donors. Data were analyzed using paired t test. Asterisks mark significant differences: *p<0.05, ns, not significant.

Figure 3

SARS-CoV-2 infection downregulates the proportion of DNAM-1⁺ and CD161⁺ NK cells due to NK cell co-culture with virus-infected HAE. Flow cytometry analysis of the proportion of NK cells expressing surface receptors CD16, NKp46, NKG2D, DNAM-1, NKG2C, CD161, NKG2A, TIM-3, TIGIT, and PD-1, in NK cells co-cultured with SARS-CoV-2 and mock-infected HAE. Data were obtained from five independent experiments, using NK cells isolated from five different healthy blood donors. Data were analyzed using paired t test. Asterisks mark significant differences: *p<0.05, ns, not significant.

Figure 4

SARS-CoV-2 infection downregulates NK cell cytotoxicity after NK cell co-culture with infected HAE. NK cell cytotoxicity assay against K562 cells. Data were obtained from four independent experiments; each experiment was performed in duplicate. Data were analyzed using t test and mean ± SEM is shown (A). Flow cytometry representative result. NK cells were co-cultured with CFSE-labelled K562 cells, in a ratio 5:1. Dead cells are DEAD Dye positive (B). Asterisks mark significant differences: *p<0.05.

Figure 5

CD161⁺ NK cells more effectively limit SARS-CoV-2 infection. Sorted CD161⁺ and CD161^- NK cells were added onto SARS-CoV-2 infected Vero (A) or A549^ACE2/TMPRSS2 (B) cells 24h p.i. Infected cells were analyzed based on the intracellular expression of SARS-CoV-2 N protein, while live Vero and A549^ACE2/TMPRSS2 cells were analyzed on FSC/SSC scatter and DEAD Dye staining. The percentage of live SARS-CoV-2 N protein⁺ cells were calculated on the basis of the percentage in G4 quadrant and the percentage of Vero (A) or A549^ACE2/TMPRSS2 (B) cells. The flow cytometry plots shows representative results. Data were obtained from five independent experiments. Data were analyzed using paired t test and means ± SEM are shown. Asterisks mark significant differences: *p<0.05, ns, not significant.

Figure 6

SARS-CoV-2 infection upregulates LLT1 expression on infected epithelial cells. Confocal microscopy images of A549^ACE2/TMPRSS2 cells (A) and HAE (B), mock or virus-infected. Images were obtained at 48h p.i. (A), or 104h p.i. (B). SARS-CoV-2 N protein is shown in red, LLT1 in green, and nuclei in blue. The bottom panels show an orthogonal view of cells (XZ projection) (B).

Figure 7

LLT1 treatment downregulates CD161⁺ NK cell proportion and impairs NK cell functions and ability to control SARS-CoV-2 infection. NK cells treated for 2h with LLT1 protein, followed by 18h stimulation with PMA/ionomycin, were analyzed for intracellular expression of granzyme B, IFN-γ, and TNF-α (A). NK cells were treated for 2h with LLT1 protein and their cytotoxicity capacity was analyzed against K562 cells (B) or their degranulation level measured as CD107a surface expression (B). The proportion of CD161⁺ NK cells was measured after 2h treatment with soluble LLT1 protein (C). SARS-CoV-2 infected cells were analyzed based on the intracellular expression of SARS-CoV-2 N protein (D). Data were obtained from five independent experiments and analyzed using paired t test. Asterisks mark significant differences: *p<0.05, ns, not significant.

Figure 8

CD161 blocking mAb pre-treatment of NK cells upregulates the proportion of granzyme B expressing NK cells and induce their cytotoxicity capacity, yet not the degranulation level. NK cells pre-treated with anti-CD161 blocking mAb (clone HP-3G10) or isotype control, and then treated for 2h with LLT1 protein, followed by 18h stimulation with PMA/ionomycin, were analyzed for intracellular expression of granzyme B, perforin, IFN-γ, and TNF-α (A). NK cells were pre-treated with anti-CD161 blocking mAb (clone HP-3G10) or isotype control, and then treated for 2h with LLT1 protein and their cytotoxicity capacity was analyzed against K562 cells (B) or their degranulation level measured as CD107a surface expression (C). FACS plots show representative results of cytotoxicity assay against CFSE-labelled K562 cells (B) or CD107a degranulation assay (C). Data were obtained from three independent experiments and analyzed using paired t test. Asterisks mark significant differences: *p<0.05, ns, not significant.

Figure 9

Scheme of SARS-CoV-2 mediated NK cell impairment by the activation of the LLT1-CD161 axis. LLT1 expression is upregulated in SARS-CoV-2 infected epithelial cells. LLT1 acts as a ligand for the NK cell CD161 receptor, which, in turn, inhibits granzyme B production by NK cells, inhibiting NK cell-mediated antiviral functions, their cytotoxicity capacity, yet not the degranulation level. Created with BioRender.com.