Table 1.
Information of subjects enrolled in scRNA-seq.
Fig 1.
ScRNA-seq analysis of peripheral blood NK cells in HFRS patients and uninfected control samples.
(A) The t-SNE presentation of the cell clusters within PBMC, T/NKT/NK cells and NK cells respectively from HFRS patients. (B)The t-SNE plots showed the signature genes (CD3D, FCER1G, NCAM1, NCR1, PTPRC and TYROBP) of NK cells in T/NKT/NK cells. (C) The t-SNE plots and (D) bar graphs of NK cells in HFRS patients with different disease severities and uninfected controls. (E) The t-SNE plots and (F) bar graphs of the proportion of NK clusters across HFRS patients with different disease severities and uninfected controls.
Fig 2.
The expression of representative genes in different subclusters of NK cells.
(A) The violin graph showing the comparison of representative genes (CD56/NCAM1, CD16/FCER1G, NKG2A/KLRC1, NKG2C/KLRC2, CD94/KLRD1 and CD57/B3GAT1) expression in each cluster of NK cells. (B) The heatmap presenting gene expression profiles of each cluster in NK cells. (C) The pathway analysis showing the pathway activation in cluster 4 and 5 of NK cells.
Fig 3.
The frequencies and phenotypic characteristics of CD56dim NK cells in HFRS patients.
(A) The representative flow cytometric plots and (B) the comparison of the frequencies of CD56dimCD16+ NK cells between patients and uninfected controls. (C) The representative histogram figures and the comparison of (D) the frequencies as well as (E) the mean fluorescence intensity (MFI) of NKG2A or NKG2C gated on CD56dimCD16+ NK cells between patients and uninfected controls respectively. (F) The comparison of the frequencies or MFI between NKG2A+ and NKG2C+ cells gated on CD56dimCD16+ NK cells in HFRS patients. (G) The representative flow cytometric graph for the expression of CD57 in CD56dimCD16+NKG2A+NKG2C- NK cells. (H) The comparison of the frequencies between CD57- and CD57+ cells gated on CD56dimCD16+NKG2A+NKG2C- NK cells of HFRS patients. Statistical analysis was performed using the Mann-Whitney U test. p-values less than 0.05 were considered statistically significant (p < 0.05, *; p < 0.01, **; p < 0.001, ***).
Fig 4.
The activation status of CD56dimCD16+NKG2A+NKG2C-CD57- NK cells in HFRS patients.
(A) Representative flow cytometric graphs and (B) the comparison of the frequencies of HLA-DR+ or CD69+ cells gated on CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among HFRS patients with different disease severities and uninfected controls. (C) Representative flow cytometric plots and (D) the comparison of the frequencies of HLA-DR+ or CD69+ cells gated on CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among HFRS patients at different stages and uninfected controls. Statistical analysis was performed using the Mann-Whitney U test. p-values below 0.05 were considered statistically significant (p < 0.01, **; p < 0.001, ***).
Fig 5.
The proliferation capacity of CD56dimCD16+NKG2A+NKG2C-CD57- NK cells in HFRS patients with different disease severity or at different stages.
(A) Representative flow cytometric plots and (B) the comparison of the frequencies of Ki67+CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among HFRS patients with different disease severity and uninfected controls. (C) Representative flow cytometric graphs and (D) the comparison of the frequencies of Ki67+CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among HFRS patients at different disease stages and uninfected controls. Statistical analysis was performed using the Mann-Whitney U test. p-values below 0.05 were considered statistically significant (p < 0.01, **; p < 0.001, ***).
Fig 6.
The frequencies of TNF-α and IFN-γ secretion of CD56dimCD16+NKG2A+NKG2C-CD57- NK cells in HFRS patients with different disease severity or at different stages.
Representative flow cytometric plots and comparison of the frequencies of TNF-α (A and B) and IFN-γ (C and D) production- CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among mild/moderate HFRS patients, severe/critical HFRS patients and uninfected controls respectively. (E) Representative flow cytometric plots and (F) comparison of the frequencies of TNF-α+IFN-γ+CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among mild/moderate HFRS patients, severe/critical HFRS patients and uninfected controls respectively. Representative flow cytometric plots and the comparison of the frequencies of IFN-γ (G and H) and TNF-α (I and J) production- CD56dimCD16+NKG2A+NKG2C-CD57- NK cells in HFRS patients at different disease stages and uninfected controls, respectively. (K) Representative flow cytometric plots and (L) comparison of the frequencies of TNF-α+IFN-γ+CD56dimCD16+NKG2A+NKG2C-CD57- NK cells in HFRS patients at different disease stages and uninfected controls, respectively. Statistical analysis was performed using the Mann-Whitney U test. The p-values less than 0.05 were considered to be statistically significant (p < 0.05, *; p < 0.01, **; p < 0.001, ***).
Fig 7.
The frequencies of cytotoxic mediators production of CD56dimCD16+NKG2A+NKG2C-CD57- NK cells in HFRS patients with different disease severity or at different stages.
Representative flow cytometric plots and the comparison of the expression of perforin (A and B), granzyme B (C and D) and CD107a (E and F) on CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among mild/moderate HFRS patients, severe/critical HFRS patients and uninfected controls, respectively. Representative flow cytometric plots and the comparison of the frequencies of perforin (G and H), granzyme B (I and J) production and CD107a (K and L) expression on CD56dimCD16+NKG2A+NKG2C-CD57- NK cells among patients at different stages of the illness and uninfected controls, respectively. The Mann-Whitney U test was used for Statistical analysis. The p-values less than 0.05 were considered to be statistically significant (p < 0.05, *; p < 0.01, **; p < 0.001, ***).
Table 2.
Screen of HTNV-derived HLA-E peptides by the K562/HLA-E*0103 cell binding assay.
Fig 8.
The screen for the HTNV-derived peptides binding to HLA-E*0103 and evaluation of the binding stability and concentration dependence.
(A) Overlay histogram showed a rightward shift with higher fluorescence intensity (FI) in the curve of K562/HLA-E*0103 cells incubated with each of the four indicated HTNV-derived peptides compared to cells incubated without peptide, indicating binding to HLA-E*0103. (B) The mean fluorescence intensity (MFI) of HLA-E expression on K562/HLA-E*0103 cells pulsed with different peptide concentrations for each of the four indicated HTNV-derived peptides or control peptides (HLA-Cw*03 or MAGE-1). (C) The MFI change tendency of HLA-E expression at the different time points on K562/HLA-E*0103 cells pulsed with each of the four indicated HTNV-derived peptides or control peptides (HLA-Cw*03 or MAGE-1) at 100uM.
Fig 9.
Detection of the binding capacity between HLA-E/HTNV peptide complexes and CD94/NKG2A(C) receptor.
(A) Representative flow cytometric plots and (B) the comparison of the frequencies of CD94/NKG2A+HLA-E+ cells after incubation with each of the four HTNV-derived peptide or control peptide. (C) Representative flow cytometric plots and (D) the comparison of the frequencies of CD94/NKG2C+HLA-E+ cells after each of the four HTNV-derived peptide or control peptide incubation. The peptide from HLA-Cw*03 leading sequence (VMAPRTLIL) served as the positive control. The peptide from MAGE-1 (EADPTGHSY) served as the negative control. Statistical analysis was performed using the Mann-Whitney U test. p-values below 0.05 were considered statistically significant (p < 0.05, *; p < 0.01, **; p < 0.001, ***).
Fig 10.
The tetramer staining to detect the binding capacity of HLA-E/HTNV peptide complex to CD56dimCD16+NKG2A+ NK cells in PBMCs from HFRS patients.
(A) Representative flow cytometric plots and (B) comparison of the frequencies of HLA-E/peptide (NP5 or GP8) tetramer+ cells gated on CD56dimCD16+NKG2A+NKG2C-CD57- NK cells between HFRS patients and uninfected controls.
Fig 11.
Activation, proliferation and effector capacities of CD56dimNKG2A+NK cells in HFRS patients by co-culture with HTNV-NP5 peptide-pulsed K562/HLA-E*0103 cells ex vivo.
Representative flow cytometric plots and changes in the frequencies of activated (HLA-DR+) (A and B), proliferated (Ki67+) (C and D) and functional (TNF-α+, perforin+ and CD107a+) (E-J) CD56dimNKG2A+NK cells between HFRS patients and uninfected controls in the presence of the HTNV-NP5 peptide. Total CD3-CD56+NK cells were isolated from the PBMCs of each HFRS patient and uninfected control using a human NK cell negative selection kit. The isolated NK cells were then co-cultured with HTNV-NP5 peptide-pulsed K562/HLA-E*0103 cells ex vivo. Statistical analysis was performed using the Mann-Whitney U test, with p-values below 0.05 considered statistically significant (p < 0.05, *; p < 0.01, **).
Fig 12.
Activation, proliferation, TNF-α and cytotoxic mediator production of CD56dimNKG2A+NK cells upon co-culture with K562/HLA-E cells pulsed with the indicated peptides ex vivo.
Representative flow cytometric (FCM) analysis and cumulative results showing the frequencies of CD107a+ (A-D), perforin+ (E-H) and TNF-α+ (I-L) in CD56dimNKG2A+ and CD56dimNKG2A-NK cells from PBMCs of uninfected controls across different peptide groups. K562/HLA-E*0103 cells loaded with peptide from HLA-Cw*03 leading sequence (VMAPRTLIL) were used as the positive control, while K562/HLA-E*0103 cells loaded with the peptide from HCMV phosphoprotein PP65 (NLVPMVATV) or no peptide were used as negative controls. Statistical analysis was performed using the Mann-Whitney U test, with p-values below 0.05 considered statistically significant (p < 0.05, *; p < 0.01, **).
Fig 13.
Ability of the HLA-E/HTNV-derived peptide complex to induce CD56dimNKG2A+NK cell-mediated lysis.
(A) Representative flow cytometric plots and (B) Kinetics and comparison showing the percentages of targeted cells killed by CD56dimNKG2A+ and CD56dimNKG2A- NK cells at effector-to-target cell ratios of 3:1 in different peptides groups. EGFP-labeled K562/HLA-E*0103 cells loaded with indicated peptide were used as target cells, while FACS-sorted CD3-CD56dimNKG2A+ and CD3-CD56dimNKG2A- NK cells from the PBMCs of uninfected controls were used as effector cells. K562/HLA-E*0103 cells loaded with the peptide from the HLA-Cw*03 leading sequence (VMAPRTLIL) served as the positive control, whereas K562/HLA-E0103 cells loaded with the peptide from HCMV phosphoprotein PP65 (NLVPMVATV) or no peptide served as negative controls. Fixable Viability Dye eFluor 450 was used to label dead cells.
Fig 14.
Proposed model illustrating the interaction between HLA-E/peptide complexes and NKG2A on NK cells.
Normally, HLA-E presented peptides derived from the leading sequence of classical HLA class I (HLA-Ia) and HLA-G molecules. Therefore, NKG2A could inhibit the cytotoxicity of NK cells to autologous healthy cells through interacting with the HLA-E/peptide complex (left). During HTNV infection, HLA-E formed a complex with HTNV-derived peptides. The HLA-E/HTNV peptide complex was unable to interact with NKG2A, leading to the enhanced antiviral activity of NK cells (right).