Single-cell landscape revealed immune characteristics associated with disease phases in brucellosis patients

Abstract

A comprehensive immune landscape for Brucella infection is crucial for developing new treatments for brucellosis. Here, we utilized single-cell RNA sequencing (scRNA-seq) of 290,369 cells from 35 individuals, including 29 brucellosis patients from acute (n = 10), sub-acute (n = 9), and chronic (n = 10) phases as well as six healthy donors. Enzyme-linked immunosorbent assays were applied for validation within this cohort. Brucella infection caused a significant change in the composition of peripheral immune cells and inflammation was a key feature of brucellosis. Acute patients are characterized by potential cytokine storms resulting from systemic upregulation of S100A8/A9, primarily due to classical monocytes. Cytokine storm may be mediated by activating S100A8/A9-TLR4-MyD88 signaling pathway. Moreover, monocytic myeloid-derived suppressor cells were the probable contributors to immune paralysis in acute patients. Chronic patients are characterized by a dysregulated Th1 response, marked by reduced expression of IFN-γ and Th1 signatures as well as a high exhausted state. Additionally, Brucella infection can suppress apoptosis in myeloid cells (e.g., mDCs, classical monocytes), inhibit antigen presentation in professional antigen-presenting cells (APCs; e.g., mDC) and nonprofessional APCs (e.g., monocytes), and induce exhaustion in CD8⁺ T/NK cells, potentially resulting in the establishment of chronic infection. Overall, our study systemically deciphered the coordinated immune responses of Brucella at different phases of the infection, which facilitated a full understanding of the immunopathogenesis of brucellosis and may aid the development of new effective therapeutic strategies, especially for those with chronic infection.

Keywords: Brucella infection; brucellosis; cytokine storm; immune response; single‐cell sequencing.

Figure 1

An overview of the results and study design for our peripheral blood mononuclear cell (PBMC) single‐cell transcriptomic study. (A) Diagram outlining the overall study design. Thirty‐five samples were obtained from 35 individuals, including 29 brucellosis patients (10 patients in acute phase, nine patients in sub‐acute phase, and 10 patients in chronic phase) and six healthy donors. (B) Box plots illustrating the log₁₀ transformed number of cells for each sample. (C) The clustering result (left row) of the nine major cell types (right row) from 35 samples. Each point represents one single cell, colored according to cell type. (D) Dot plots of the nine major cell types (columns) and expression of their marker genes (rows). (E) Disease preference of major cell clusters as estimated using RO/E. (F) Heatmap showing the association between cell composition and disease types. The color represents analysis of variance (ANOVA) q values.

Figure 2

Associations between brucellosis disease phase and peripheral blood mononuclear cell (PBMC) cellular composition. (A) UMAP projection showing the 32 cellular subtypes identified from 35 samples. Each dot depicts a single cell while the color represents the cell subtype. (B) Dot plot depicting the disease preference for each of the 32 cell subtype as calculated using RO/E. (C) Heatmap showing the p values from analysis of variance (ANOVA) of differences in cell subtype composition between disease phases. Disease phase: HD, AC, SA, and CH. (D) Classes of heavy chains for plasma cells from brucellosis patients. (E) Bar plots (left) showing IFNG expression in CD4_Th1 cells between different groups, Box plots (right) showing plasma level of IFN‐γ across different phases. (F) Bar plot showing Th1 signature expression in CD4_Th1 cells between different groups. (G) PAGA analysis of CD8⁺ T cell pseudo‐time: the associated cell type and the corresponding status are listed. (H) Bar plots showing S100A8/A9/A12 and HLA‐DRA/B1/B5 expression in Mono_MDSCs between different groups. (I) Pie chart depicting the relative contribution of each cell subtype to the C1 complement components. (J) UMAP projection density plots of Mono_CD14 cells from different groups.

Figure 3

Contribution of S100A8/A9 to potential cytokine storms in acute patients. (A) Bar plots showing cytokine scores and inflammatory scores across different groups. (B) Uniform manifold approximation and projection (UMAP) projections of peripheral blood mononuclear cells (PBMCs). Colored based on the nine major cell types (top left), seven hyper‐inflammatory cell subtypes (top right), cytokine (middle) and inflammatory score (bottom). (C) Pie charts depicting the relative contribution of each inflammatory cell subtype to the cytokine and inflammatory scores. (D) Heatmap depicting the expression of cytokines within each hyper‐inflammatory cell subtype identified. (E) Bar chart depicting the relative contribution of the top 10 cytokines in patients with acute brucellosis. (F) Box plots showing S100A8 and S100A9 expression across different groups. (G) Box plots showing plasma profiling of S100A8/A9 complex across different groups. (H) The expression analysis of S100A8/A9‐TLR4‐MyD88 pathway. (I) Heatmap depicting the expression of selected genes across different groups. (J) Heatmap of the sum of significant interaction among the seven hyper‐inflammatory cell subtypes. (K) Circos plot depicting the ligand‐receptor pair interactions between Mono_CD14 and the seven hyper‐inflammatory cell subtypes.

Figure 4

Immunological features of CD4⁺ T‐cell subsets. (A) Box plot showing the exhaustion score in CD4_Th1 cell subtype across different groups. (B) Dot plot showing the expression of selected exhaustion genes in CD4_Th1 cell subtype across different groups. (C) Box plot showing the relative percentage (left), regulatory scores (middle) and TGF‐β expression of CD4_Treg cell subtype across different groups. (D) Box plot showing the apoptosis and migration scores of CD4⁺ T cells from different groups. (E) Heatmap depicting the expression of apoptosis‐related genes across different groups. (F) Venn diagram (left) illustrating the number of upregulated genes in CD4⁺ T cells and box plots (right) of shared GO terms of CD4⁺ T cells across different conditions. (G) Box plot showing the indicated functional scores (IFN‐γ response scores) of CD4⁺ T cells.

Figure 5

Immunological features of CD8⁺ T‐cell subsets. (A) Top upregulated genes for each CD8⁺ T‐cell cluster was calculated, and genes with high frequencies are displayed. (B) Box plot showing the exhaustion scores in CD8⁺ T cells across different groups. (C) Box plots showing the exhaustion score in effector CD8⁺ T cells from different groups. (D) Heatmap depicting the expression of exhaustion‐related genes in effector CD8⁺ T cells. (E) Uniform manifold approximation and projection (UMAP) projections for cytotoxic scores in CD8⁺ T cells across different conditions. (F) Box plot showing the apoptosis scores in CD8⁺ T cells across different groups. (G) Heatmap depicting the expression of apoptosis‐related genes in CD8⁺ T cells across different groups. (H) Venn diagram (left) illustrating the number of upregulated genes in CD8⁺ T cells and box plots (right) of shared GO terms of CD8⁺ T cells across different conditions. (I) Box plot showing the IFN‐γ response scores of CD8⁺ T cells from different cells.

Figure 6

Immunological features of NK‐cell subsets. (A) Heatmap depicting the expression of activation‐related genes in effector NK cells across different conditions. (B) Bar plots showing the IFN‐γ expression in NK_CD56^(Bri) cells between different groups. (C) Box plot showing the cytotoxic scores in NK_CD56^(Dim) cells across different groups. (D) Heatmap depicting the expression of cytotoxicity‐related genes in NK_CD56^(Dim) cells across different conditions. (E) Box plot showing the exhaustion score in NK cells across different groups. (F) Heatmap depicting the expression of exhaustion‐related genes in NK_CD56^(Dim) cells across different conditions. (G) Box plot (left) showing the migration score in NK cells across different groups, heatmap (right) depicting the expression of migration‐related genes in NK cells across different conditions. (H) Box plot (left) showing the apoptosis score in NK cells across different groups, heatmap (right) depicting the expression of apoptosis‐related genes in NK cells across different conditions. (I) Heatmap depicting the expression of selected genes in NK cells across different conditions.

Figure 7

Immunological features of myeloid subsets in brucellosis patients. (A) Box plot showing the phagocytosis (left), antigen presentation (middle) and apoptosis scores (right) in mDCs across different groups. (B) Venn diagram illustrating the number of upregulated genes (left) and downregulated genes in Mono_MDSCs. (C) Violin plots showing the expression of HLA‐DRA/B5/B1 across monocyte subsets. (D) Venn diagram illustrating the number of upregulated genes in classical monocytes (left), selected enriched GO terms (right) for genes upregulated in classical monocytes. (E) Box plot showing the IFN response in clinical monocytes across different groups. (F) Violin plots showing expression of typical inflammatory cytokines in clinical monocytes across different groups. (G) Box plots showing plasma profiling of MIP‐α (CCL3), MIP‐1‐β (CCL4), IL1‐β (IL‐1B), TNF‐α (TNF) and IL‐8 (CXCL8) across different groups. (H) Violin plots showing expression of PIM1, PIM3 and KLF4 in clinical monocytes across different groups.