Figure 1.
Runx3 expression is restricted to splenic CD4+/CD11b+ DCs and to fraction of CMP and MDP.
(A) Schematic representation of the Runx3 gene and the targeted alleles used to generate the Runx3-GFP reporter mice. P1 and P2 represents promoter 1 and 2 respectively. (B) Flow cytometry analysis of WT or Runx3P1/P2-GFP BM cells gated for lineage negative cells and stained for myeloid progenitor markers to identify GFP expression in CMP, MDP, CDP and preDC. WT littermate and Runx3P1/P2-GFP mice are marked by red and blue lines, respectively. (C) Flow cytometry analysis of cells isolated from WT or Runx3P1/P2-GFP mice. Dot plot of isolated splenic CD11c+MHCII+ DC stained for CD4+ and CD8+ DC subsets (upper panel). CD4+ and DN DC express Runx3 (lower panel histograms). CD8+ DC are GFP negative indicating lack of Runx3 expression. All data in this figure are representative of three independent experiments with one to two mice per group.
Figure 2.
Runx3 is required for splenic CD11b+ DC lineage development.
(A) Representative profiles of splenocytes isolated from Cebpα-DC-Runx3Δ (left) or CD11c-DC-Runx3Δ (right) mice stained for CD11c and MHCII. Comparisons between Cre- (upper) and Cre+ (lower) littermates are shown. (B) Representative profiles of gated CD11chiMHCII+ splenic DC isolated from Cebpα-DC-Runx3Δ (left) or CD11c-DC-Runx3Δ (right), stained for CD4 and CD8. Comparisons between Cre- (upper) and Cre+ (lower) littermates are shown. Note the reduced proportion of CD4+ DC upon loss of Runx3. (C and D) Splenocytes were isolated from Cebpα-DC-Runx3Δ (C), CD11c-DC-Runx3Δ (D) or littermate control WT mice (n=4, 6-8 week old). Cells were counted and total cell numbers per mg tissue calculated and compared to that of littermate control. Each dot represents one independent animal. *P<0.05, **P<0.01, ***P<0.001 (Students two-tailed t test). Shown are results of one experiment out of two with the same finding. (E) Representative profiles of splenocytes isolated from Cebpα-DC-Runx3Δ (left) or CD11c-DC-Runx3Δ (right) mice gated for CD11chiMHCII+ DC subset and analyzed for Esam and CD8. Comparisons between Cre- (upper) and Cre+ (lower) littermates are shown. Shown representative of three independent experiments with one to two mice per group. (F) Lack of apoptosis among Runx3-deficient cDC. Representative profiles of Runx3+/+ (left panel) and Runx3-/- (right panel) cDCs stained with Annexin-V and with 7AAD viability dye. Shown representative of two independent experiments with one to two mice per group. (G) Graphic summary of competitive BM-repopulation assay. Flow cytometric analysis of [Runx3+/+ (CD45.2) /WT (CD45.1) > WT (CD45.1)] (left) and [CD11c-DC-Runx3Δ (CD45.2) /WT (CD45.1) > WT (CD45.1)] (right) chimeric mice 6-8 weeks after transplantation. CD45.2/CD45.1 ratios were calculated for each cell population. Values <1 indicate out-competition of the mutant by WT (CD45.1) cells, whereas values >1 show an advantage of Runx3Δ (CD45.2) cells. Representative results from one out of two independent experiments are shown (mean ± SD) with 3 or more animals in each group. *P<0.05, **P<0.01, ***P<0.001 (Students two-tailed t test).
Figure 3.
Alteration in CD4+ and Esamhi DC gene expression due to loss of Runx3.
(A) Scatter plot of differentially expressed genes in Runx3-deficient CD4+ DC and WT control DC. Splenic CD4+ DC from three Runx3Δ and WT mice were isolated by FACS, RNA was isolated and subjected to microarray analysis. Shown are the mean intensities (log2) of WT and Runx3Δ CD4+ DC. Genes that were up- or down- regulated due to Runx3Δ are marked by red or blue, respectively. Differential expression cut-off was set to minimal absolute fold-change of 1.5, and p-value<=0.05. Selected relevant genes are indicated. (B) Expression of Esam (upper panel) CD11b (middle panel) and CD11c (lower panel) in splenic DC subsets from Runx3+/+ (Runx3fl/fl) mice (red line) and CD11c-DC-Runx3Δ mice (blue line). (C) Scatter plot of differentially expressed genes in CD4+Esamhi DC from WT and CD11c-DC-Runx3Δ mice. Splenic Esamhi DC were obtained from two individual mice. Genes that were up- or down- regulated due to Runx3Δ are marked by red or blue, respectively. Differential expression cut-off was set to minimal absolute fold-change of 1.5, and p-value<=0.05. Selected relevant genes are indicated. (D) Gene expression log2 pairwise comparisons between WT-CD4+Esamhi, CD11c-Runx3ΔCD4+Esamhi subsets and the published Esam experiment [8]. Compared are 162 genes differentially expressed in the DC-Runx3ΔEsamhi experiment and in Lewis et al, 2011 Esam experiment [8]. The shown comparison of gene expression intensities indicates Spearman correlation values numerically and by color (above main diagonal), histograms of expression values (main diagonal) and scatter plots (below main diagonal). (E) Clusters showing Log2 ratio of gene expression of 224 genes (see Methods) correlated with Runx3 occupied genomic regions. Runx3 binding (up to a distance of 200kb from TSS) is presented as column No-1. Gene expression (Log2 ratio) was calculated for the following combination: column No-2 Runx3-/-CD4+ vs. WT-CD4+, column No-3 DC-Runx3ΔEsamhi vs. CD4+, column No-4 WT Esamlow vs. Esamhi and column No-5 DC-RbpjΔ vs. Esamhi. Levels of up or down regulation are color-coded. (F) RT-qPCR analysis of genes scored by differential gene expression of WT Esamhi and Esamlow CD11b vs. Runx3Δ Esamhi populations. Data represents normalized expression values relative to WT Esamhi sample (±SD of three assays from two biological repeats). (G) Histograms for Kit and Flt3 expression in Esamhi DC subset. Red and blue lines represents WT CD11c (Runx3fl/fl) and CD11c-Runx3Δ mice, respectively.
Figure 4.
Analysis of Runx3 occupied genomic regions in splenic CD4+ DC.
(A) Venn diagram summarizing the overlap between Runx3 ChIP-seq bound genes and differentially expressed genes in splenic CD4+ DC. (B) Runx3 binding-pattern on five genomic loci differentially expressed in CD4+ DC. Shown are ChIP-seq Runx3 peaks and the normalized read coverage of whole cell extract (top) and Runx3 ChIP (bottom) uploaded to UCSC Genome Browser mm9 genome assembly. The position of MACS2-identifird Runx3 peaks is indicated by rectangles. For each panel the scale bar and diagram of gene exon structure are presented on top or bottom, respectively. (C) Enriched RUNX and zDC TF motif among Runx3 bound regions. Motifs were identified de-novo using the MEME-ChIP software.
Figure 5.
Loss of Runx3 abrogates functional activities of CD4+ DC.
(A) CD4+ T cell priming in DC-Runx3Δ mice in vivo. CFSE-labeled OVA-specific OT-I CD8+ T cells or OT-II CD4+ T cells were introduced to DC-Runx3Δ or WT control littermate mice followed by immunization with OVA. Shown are flow cytometric analyses of transgenic T cell grafts recovered from immunized recipient mice. Proliferation of OT-I CD8+ T cells (top) and OT-II CD4+ T cells (bottom) cells in DC-Runx3+/+ (red) or DC-Runx3Δ mice (blue) compared to PBS immunized mice (green) and non-labeled cells (brown). Shown representative of two independent experiments with two to three mice per group. (B) Total cell number of OT-II CD4+ T cell Proliferation quantification. PBS-immunized mice T cell proliferation is set as one (n=5). ***P < 0.001 (Student’s two-tailed t test). (C) Runx3 binding-pattern on MHCII genomic loci. Shown are ChIP-Seq tracing wiggle files uploaded to UCSC Genome Browser mm9 genome assembly and the normalized read coverage of whole cell extract (top) and Runx3 ChIP (bottom). The position of MACS2-identified Runx3 peaks is indicated by rectangles. For each panel the scale bar and diagram of gene exon structure are presented on top or bottom, respectively. (D and E) The capacity of splenic DC subtypes to phagocytose latex beads was analyzed in DC-Runx3Δ (blue) and WT (red) littermate mice 18h after i.v. injection of 1010 FITC-labeled 0.5μm latex beads. Shown are histograms of FITC-fluorescence in CD4+, CD8+ and DN DC subsets (D) or in Esamhi and Esamlow DC subsets (E). Shown representative of two independent experiments with two mice per group with the same finding. (F) The capacity of splenic DC to phagocytose latex beads was further analyzed using mixed BM chimera mice 18h after i.v. injection of 1010 FITC-labeled 0.5μm latex beads. Shown are histograms of FITC-fluorescence in CD4+ and Esamhi DC subsets gated on WT CD45.1 (red) or DC-Runx3Δ CD45.2 (red). Shown representative of two independent experiments with two mice per group.