Chemokine expression profile of an innate granuloma

Megan E Amason
Cole J Beatty
Carissa K Harvest
Daniel R Saban
Edward A Miao author has email address

Department of Integrative Immunobiology, Duke University School of Medicine, Durham, NC, USA 27710
Department of Molecular Genetics and Microbiology, Duke University School of Medicine, Durham, NC, USA 27710
Department of Microbiology and Immunology, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 27599
Department of Pathology, Duke University School of Medicine, Durham, NC, USA 27710
Department of Cell Biology, Duke University School of Medicine, Durham, NC, USA 27710
Department of Ophthalmology, Duke University School of Medicine, Durham, NC, USA

https://doi.org/10.7554/eLife.96425.2

Open access
Copyright information

Figures and data

Spatial transcriptomics dataset reveals 16 unique clusters during infection with C. violaceum.
(A) SpatialDimPlots showing hematoxylin and eosin (H&E) and cluster overlay of spatial transcriptomics data corresponding to various days post-infection (DPI). Each circle is an individual barcoded spot that is 55 µm in diameter. (B) UMAP plot of 16 unique clusters identified based on differentially expressed genes during the course of infection. Characterization of predominant cell types and/or location of each cluster (initial characterization performed in Harvest et al., 2023); Macrophage zone (M), hepatocyte (HEP), representative HEP (rep HEP), necrotic core center (NC-C), NC-periphery (NC-P), coagulative necrosis (CN), CN-macrophage (CN-M), endothelial cell (EC), outside granuloma (OG). (C) Temporal prevalence of CD45⁺ clusters, calculated as proportion of spots represented by each cluster within each timepoint. (D) SpatialDimPlot at 10 DPI as in (A), showing cluster overlay and annotated with cluster identity. (E) SpatialFeaturePlot at 10 DPI, showing log-normalized expression of Pf4 (murine homolog of CXCL4).

Expression level of chemokine ligands during infection with C. violaceum.
Expression was visually ranked as absent, low, medium, or high based on SpatialFeaturePlots. Maximum expression rank recorded here. Table generated from: (David & Kubes, 2019; Hughes & Nibbs, 2018; Sokol & Luster, 2015; Zlotnik & Yoshie, 2000, 2012). Lymph node (LN); Natural killer cell (NK); NK T cell (NKT); innate lymphoid cell (ILC); dendritic cell (DC).

Expression level of chemokine receptors during infection with C. violaceum.
Expression was visually ranked as absent, low, medium, or high based on SpatialFeaturePlots. Maximum expression rank recorded here. Table generated from: (David & Kubes, 2019; Hughes & Nibbs, 2018; Sokol & Luster, 2015; Zlotnik & Yoshie, 2000, 2012). Natural killer cell (NK); innate lymphoid cell (ILC); dendritic cell (DC); plasmacytoid DC (pDC); lymph node (LN); red blood cell (RBC).

Expression level of selected proteins and receptors during infection with C. violaceum.
Expression was visually ranked as absent, low, medium, or high based on SpatialFeaturePlots. Maximum expression rank recorded here. Table generated from: (Bui et al., 2020; David & Kubes, 2019; Parks et al., 2004; Wang et al., 2018). Dendritic cell (DC); plasmacytoid DC (pDC); Kupffer cell (KC); natural killer cell (NK); syndecan 1 (SDC1).

Top twenty differentially expressed genes per cluster.
The FindAllMarkers function was used to identify the top differentially expressed genes for each cluster across all timepoints. Genes were sorted from highest to lowest Average log2 fold change (avg_log2FC) values within each cluster. Genes of interest shown in red. Full dataset found in Table 4 – source data 1.

Chemokines involved in neutrophil recruitment are upregulated during infection.
SpatialFeaturePlots displaying normalized gene expression data of CXCR2 ligands (i.e. Cxcl1, Cxcl2, Cxcl3, and Cxcl5) at various days post-infection (DPI). Scale set at 0 – 3.0 expression.

Chemokines involved in monocyte recruitment are upregulated during infection.
SpatialFeaturePlots displaying normalized gene expression data of CCR2 ligands (i.e. Ccl2, Ccl7, and Ccl12) at various days post-infection (DPI). Scale set at 0 – 3.0 expression.

Qualitative heatmaps of chemokine and receptor expression during infection.
Normalized expression in SpatialFeaturePlots was visually ranked as absent (grey), low (blue), medium (yellow), or high (red) for (A) CXCL family chemokines, (B) CCL family chemokines, (C) CXC chemokine receptors, and (D) CC chemokine receptors. Visual rankings were based on both the intensity of expression and the relative number of spots that expressed the gene. (A-B) Scale set at 0 – 3.0 expression; (C-D) Scale set at 0 – 2.0 expression. Arrows indicate Ligand – Receptor interactions. Ligands are color-coded based on the maximum expression level reached at any time during the course of infection.

Chemokines involved in monocyte recruitment peak after chemokines involved in neutrophil recruitment.
Comparative analysis of Cxcl1 (A, C, and E) and Ccl2 (B, D, and F). (A-B) UMAP plots of 16 unique clusters showing normalized expression level of each gene. Maximum expression level set to 1.5; annotated with cluster i entity; Macrophage zone (M), hepatocyte (HEP), representative HEP (rep HEP), necrotic core center (NC-C), NC-periphery (NC-P), coagulative necrosis (CN), CN-macrophage (CN-M), endothelial cell (EC), outside granuloma (OG). (C-D) Violin plots of 16 unique clusters showing normalized expression level of each gene across all timepoints. (E-F) Violin plots of various days post-infection (DPI) showing normalized expression level of each gene within all clusters.

CCR2 and monocyte recruitment are essential for a successful granuloma response to C. violaceum.
Wildtype (WT) and Ccr2^−/− mice were infected intraperitoneally (IP) with 1×10⁴ CFU C. violaceum. (A) Survival analysis of WT (N = 10) and Ccr2^−/− (N = 9) mice. Two experiments combined. Mantel-Cox test, ****p<0.0001. (B-K) Livers and spleens were harvested 5 days post-infection (DPI). Bacterial burdens in the (B) liver and (C) spleen of WT and Ccr2^−/− mice. Two experiments combined. Each dot represents one mouse. (B) Two-tailed t test (normally distributed data); ***p=0.0002. (C) Mann-Whitney (abnormally distributed data); **p=0.0012. Dotted line, limit of detection. Solid line, median. (D) Gross images of WT and Ccr2^−/− livers 5 DPI. (E) Gating strategy for analysis of neutrophil (Ly6G⁺) and macrophage (CD68⁺) numbers via flow cytometry. Liver samples from infected mice shown. Frequency of CD68⁺ macrophages from single cell gate in the (F) liver, (H) spleen, and (J) blood. Frequency of Ly6G⁺ neutrophils from single cell gate in the (G) liver, (I) spleen, and (K) blood. (F-K) Three experiments combined using only female mice. Each dot represents one mouse, with 10,000 events collected per sample. Two-way ANOVA (for multiple comparisons to assess genotype and infection); key comparisons and p-values shown. Line represents mean ± standard deviation.

Loss of CCR2-dependent monocyte trafficking results in abnormal granuloma architecture and failure of bacterial containment.
WT and Ccr2^−/− mice were infected intraperitoneally (IP) with 1×10⁴ CFU C. violaceum and livers were harvested 5 days post-infection (DPI). Serial sections of livers stained by hematoxylin and eosin (H&E) or various IHC markers for (A-D) WT female and (E-H) Ccr2^−/− male. Necrotic core (NC), coagulative necrosis zone (NC), macrophage zone (M). For 10X, scale bar is 100 µm. For 20X and 40X, scale bar is 50 µm. Representative of two experiments with 2 – 4 mice per group, and multiple granulomas per section.

Sign up for email alerts