Regulatory genes implicated in neutrophil activation, with further focus on IRF family members.

<p><b>A.</b> Genes were placed into 25 clusters (1–128 genes each, shown as the column headings; A and B are used to identify clusters that have the same numbers of genes) based on patterns of expression in individual samples of neutrophils from blood, SF, UA, and TG, as shown in the heatmap at the top. Clusters that clearly represented up-regulated (U) or down-regulated (D) genes (relative to blood) were pooled and were used to generate a list of predicted regulatory genes (rows) showing enrichment based on the ImmGen regulatory model. Association of each of the 64 regulators with each of the 25 gene clusters was then quantified (P-value of chi-square test), and this matrix of P-values was subjected to hierarchical clustering in order to identify related regulators (rows) and related gene clusters (columns). The lower heatmap indicates these P-values (darker = lower), and the dendrogram and colored bars on the right show groups of regulators with similar patterns of association with various gene clusters. The presence of patterns in the top heatmap (e.g., clustering of clusters characterized by up-regulation in TG, SF, or UA, or by down-regulation in SF), which shows normalized average expression in the 4 neutrophil populations in each cluster, validates this method. The group of regulators shown in light blue was associated with gene clusters indicated in bold; inspection of genes in these clusters led to implication of the type 1 interferon pathway and Irf9. <b>B.</b> Up-regulation of genes induced by type 1 interferons via Irf9, in TG and/or UA but not SF neutrophils. The heatmap shows mean expression in blood (BL), SF, UA, and TG neutrophils. Mean expression across all four conditions was placed at the center of the gradient (white) for each gene. The full color gradient for each gene represents an 8-fold difference in expression. The list of genes of interest and the pathway diagram were generated using the KEGG and Ingenuity databases. Only genes showing at least 1.5-fold differences in expression comparing conditions are shown in the heatmap. In the pathway diagram, genes showing statistically significant (Q<0.05 by ANOVA) differences that varied 2-fold in at least one pairwise comparison of conditions are shown in red, and genes showing fold differences of 1.5–2 and/or not meeting statistical significance are shown in pink. <b>C.</b><i>Irf5</i> is required for production of several cytokines and chemokines by mouse neutrophils stimulated in vitro with the TLR9 ligand CpG-B, but not for production induced by the TLR2 ligand Pam3Cys nor the TLR4 ligand LPS. The panels show the mean ± SEM of 3 independent experiments using FACS-sorted Gr1^hiCD11b⁺F4/80⁻ neutrophils. Since secretion varied between experiments but reliably did so in parallel for the different analytes, data were analyzed by determining the fold difference between <i>Irf5−/−</i> and WT in each experiment and applying one-sample T-tests to the fold-differences for the 3 experiments. P values for cells treated with CpG were 0.014 for TNF and <0.01 for the other proteins, and 0.13–0.97 for other TLR ligands.