Table 1.
Baseline characteristics of SFTS patients.
Fig 1.
The immune landscape of circulating NK cells in SFTS patients.
(A) UMAP plot showing NK cell subpopulations identified from peripheral blood of SFTS patients and healthy controls. (B) Stacked bar plot illustrating the distribution of NK cell clusters across individual samples. (C) Heatmap displaying representative marker genes defining each NK cell cluster. (D) Pseudotime trajectory analysis of NK cell differentiation states across clusters. (E) RNA velocity analysis indicating the directionality and dynamics of NK cell state transitions. (F) Heatmap of NK cell clusters categorized based on canonical functional markers. (G) Expression heatmap of activating and inhibitory NK cell receptors across identified clusters.
Fig 2.
Upregulation of TIM-3 expression in circulating NK cells from SFTS patients.
(A) Proportions and expression levels of immune checkpoint molecules in circulating NK cells. (B) Volcano plot showing differentially expressed genes between TIM-3+ and TIM-3- NK cells. (C) Enriched pathways identified by Metascape comparing TIM-3+ and TIM-3- NK cells. (D) Expression levels of HAVCR2 (encoding TIM-3) and its ligand genes (CEACAM1, HMGB1, LGALS9) across NK cell subsets. (E) Feature plots displaying the expression of HAVCR2 and LGALS9 in NK cells. (F) Cell-cell communication analysis by CellChat showing enhanced interactions between NK cells and other immune populations in fatal SFTS compared to recovered cases. (G) Chord diagram illustrating LGALS9-associated signaling pathways enriched in NK cells from fatal cases, including the LGALS9–HAVCR2 axis. Box plots show the median and interquartile range. Statistical analysis was performed using the Wilcoxon rank-sum test. DC, dendritic cell; NK, natural killer cell; SFTS, severe fever with thrombocytopenia syndrome.
Fig 3.
WGCNA identifies prognosis-associated hub genes in SFTS.
(A) Gene clustering dendrogram constructed by weighted gene co-expression network analysis (WGCNA). (B) Module–trait relationships showing the turquoise module as significantly correlated with clinical outcome. (C) Correlation between gene significance and module membership within the turquoise module for the fatal outcome phenotype. (D) Top 10 hub genes from the most highly connected cluster, forming an interferon-related co-expression network. (E) Top 10 hub genes from the second highly connected cluster, including HAVCR2, within a prognosis-associated regulatory network. (F) Gene Ontology (GO) enrichment analysis of the first cluster hub genes reveals immune response and interferon signaling pathways. (G) GO enrichment analysis of the second cluster hub genes highlights cytotoxic pathways.
Fig 4.
TIM-3 expression on peripheral NK cells in healthy individuals and SFTS patients.
(A) Representative flow cytometry plots showing TIM-3 expression on NK cells from healthy donors and SFTS patients. (B) Frequency of NK cells (CD3 ⁻ CD56⁺) among peripheral blood mononuclear cells (PBMCs) in healthy donors (N = 14), recovered SFTS patients (N = 17), and deceased SFTS patients (N = 4). (C) Frequency of TIM-3 ⁺ NK cells among total NK cells in the same cohorts. (D) Paired longitudinal analysis of peripheral TIM-3 + NK cell frequencies among NK cells in SFTS patients (N = 4) during the acute phase and approximately two weeks into recovery. Paired longitudinal analysis of soluble TIM-3 (E), Galectin-9 (F), TNF-α (G), IFN-γ (H) in plasma of SFTS patients (N = 9) during the acute and recovery phase. Each dot represents an individual donor. Box plots show the full range (min to max). Statistical analysis was performed using Kruskal–Wallis H test for (B); one-way ANOVA with Tukey’s multiple comparisons test for (C) and paired t-test for (D-H). IFN-γ, interferon gamma; NK, natural killer; sGalectin-9, soluble Galectin-9; SFTS, severe fever with thrombocytopenia syndrome; TNF-α, tumor necrosis factor alpha.
Fig 5.
Enhanced cytotoxic potential of TIM-3+ NK cells in SFTS patients.
(A) Representative flow cytometry plots showing granzyme B expression in TIM-3+ and TIM-3- NK cells isolated from peripheral blood of SFTS patients at disease onset and two weeks into recovery. (B-D) Quantification of granzyme B (B), perforin (C) and CD107a (D) expression in TIM-3+ and TIM-3- NK cells at disease onset (N = 3, 3, 6, respectively); median values are indicated. For CD107a detection, PBMCs of SFTS patients at disease onset were stimulated with PMA and ionomycin for 6 hours. (E) Paired longitudinal analysis of granzyme B expression in TIM-3+ and TIM-3- NK cells from matched samples at disease onset and recovery phase (N = 3). Statistical analysis was performed using paired t-tests. NK, natural killer; PMA, phorbol 12-myristate 13-acetate; SFTS, severe fever with thrombocytopenia syndrome.
Fig 6.
TIM-3 blockade attenuates cytotoxic function of NK cells in SFTS.
(A-C) Quantification of granzyme B (A), perforin (B) and CD107a (C) expression in NK cells with and without αTIM-3 antibody treatment after PMA/Ionomycin stimulation (N = 6); median values are indicated. (D) Flow cytometric analysis of IFN-γ expression in NK cells following αTIM-3 antibody blockade after PMA/Ionomycin stimulation. (E) Quantification of IFN-γ expression in NK cells with or without αTIM-3 antibody treatment (N = 6); box plots show minimum to maximum values. (F) Quantification of IFN-γ concentration in NK cells supernatant with and without αTIM-3 antibody treatment after stimulation (N = 6). (G) Flow cytometric analysis of TNF-α expression in NK cells following TIM-3 blockade after PMA/Ionomycin stimulation. (H) Quantification of TNF-α expression in NK cells with or without αTIM-3 antibody treatment (N = 6); box plots show minimum to maximum values. (I) Cytokine concentration in NK cells supernatant with and without αTIM-3 antibody treatment after stimulation using a CBA kit (N = 6); data are presented as mean ± SD. Statistical analysis was performed using paired t-test. αTIM-3 Ab, anti-TIM-3 antibody; CBA, cytometric bead array; IFN-γ, interferon gamma; MFI, mean fluorescence intensity; SD, standard deviation; TNF-α, tumor necrosis factor alpha.
Fig 7.
Soluble TIM-3 and Galectin-9 as serum biomarkers for SFTS prognosis.
(A) Serum levels of soluble TIM-3 (sTIM-3) and Galectin-9 in healthy individuals (N = 10), recovered SFTS patients (N = 17), and deceased SFTS patients (N = 4). (B) Pearson correlation analysis between sTIM-3 levels, Galectin-9 levels, and the frequency of TIM-3+ NK cells among NK cells with clinical characteristics. (C) ROC curves of serum Galectin-9, sTIM-3, and the frequency of TIM-3 ⁺ NK cells in predicting fatal outcome in the same cohort. (D) ROC curves of serum Galectin-9 and sTIM-3 in the validation cohort (N = 104), with AUC, sensitivity, and specificity calculated using the Youden index. (E) Kaplan–Meier survival analysis stratified by high vs. low serum Galectin-9 levels in the validation cohort (N = 104), using the ROC-derived optimal cutoff. Statistical significance was assessed using the Mann-Whitney test for (A), and Pearson correlation analysis with the Benjamini-Hochberg correction for (B), DeLong’s test for (C-D) and the log-rank test for (E). sTIM-3, soluble T cell immunoglobulin and mucin domain-3; Gal-9, galectin-9; CCI, Charlson comorbidity index; PLT, platelet count; AST, aspartate aminotransferase; Cr, creatinine; ROC, receiver operating characteristic curve; AUC, area under the ROC curve.