Figure 2

Direct comparison between high-throughput genomic studies hints at a complex interplay between genes as the basis for the differences between monocyte subsets. (a) The overlap between differentially expressed genes identified by Zhao et al. [6] and Ancuta et al. [7] was determined after eliminating all genes without gene symbols or duplicated. Merging the two studies we find that the number of genes differentially expressed between monocyte subsets amounts to 752 (100%). Of these 24% (145 genes) are genes detected in both studies; 48% of the remaining genes are detected as differentially expressed by [6] and 28% by [7]. A scatter-plot of the fold expression difference in CD16⁺ compared with CD16^- monocytes shows correlation of the values in the two studies. Red and green cutoff lines extend along the values 1.5 and -1.5. The upper right quadrant and the lower left quadrant show genes with similar fold differences between the studies. (b) Overlap between the monocyte subset gene lists and a dataset of genes involved in human monocyte maturation induced by MCSF and macrophage activation induced by a combination of LPS and IFN-γ (M1) or IL-4 (M2) [8]. This comparison shows that 190 out of the 434 genes selected from [6], and 180 out of the 318 genes selected from [7], are also contained in the maturation/activation gene set. The overlap of the three lists amounts to 86 genes. A hierarchical clustering of these 86 genes shows that a proportion of them are regulated by MCSF and contrastingly regulated by the combination of IFN-γ and LPS used in [8] to drive macrophages towards classical activation. The differences in expression between subsets only partially correlate with the MCSF gene expression pattern. The tree can be divided in three main clusters (right). Cluster 1 shows genes downregulated by MCSF stimulation in total monocytes and whose levels are lower in CD16⁺ monocytes. Cluster 2 shows genes induced by MCSF stimulation that are highly expressed by CD16⁺ monocytes. These two clusters support the hypothesis that part of the differences between subsets correlates with an MCSF responsive phenotype for CD16⁺ monocytes. However, the behaviors of the genes in cluster 3 do not correlate with those of MCSF stimulation, and instead correlates with the inflammatory profile induced by the combination of IFN-γ and LPS used in [8] to drive macrophages towards classical activation. A list of the 86 overlapping genes is given in Additional data file 2. Other abbreviations used: M3, macrophage; Mo, monocytes; 3 DM, Monocytes stimulated for three days with MCSF equivalent to 3^rd day macrophages; 7 DM, Monocytes stimulated for seven days with MCSF equivalent to 7^th day macrophages. In the figure the ratio of RNA expression between given categories is indicated by the forward slash or stroke symbol "/". A-CD16⁺/CD16^-, Ratio of RNA expression levels of CD16⁺ vs CD16^- monocytes provided by Ancuta et al.[7]; Z-CD16⁺/CD16^-, Ratio of CD16⁺ vs CD16^- monocytes provided by Zhao et al.[6].