Genome-wide characterisation of clinical kinase inhibitors in macrophage activation.

The possibility to control kinase activities through the use of chemical inhibitors is a key strategy for the treatment of numerous pathologies. To date, 243 kinase inhibitors are in clinical trials and of these 62 are approved by the FDA. Frequently, kinase inhibitors are designed on the basis of conserved features of the kinase domain. Although this approach proved to be highly effective, it also leads to the common occurrence of off-target effects. So far, the impact of kinase inhibitors off-target effects in vivo has not been systematically characterized. Recently, technological advances allowed the in vitro characterization of all the 243 kinase inhibitors currently in clinical trials, thus allowing defining their spectrum of selectivity. Nevertheless, the spectrum of action of these inhibitors in the context of intact signalling pathways in cells is not known. The strategy adopted to characterize kinase inhibitors in intact cells involved the identification of a specific cell type with dedicated signalling pathways and an interpretable readout which could capture all signalling changes. Mining hundreds of public gene expression datasets showed that macrophages display a most distinctive profile of expression of protein kinases. Macrophages are key cells of the innate immune system capable to respond to a wide range of stimuli with evident phenotypic changes, such as the polarization into pro-inflammatory or anti-inflammatory cells in response to lipopolysaccharide (LPS) and interleukin-4 (IL-4), respectively. I used genome-wide histone H3 lysine 27 acetylation (H3K27ac) changes as readout of the effects of selected kinase inhibitors on macrophage responses to IL-4 or LPS. As the acetylation state of genomic regulatory regions is controlled by signalling events modulating transcription factor activity, the perturbation of a signalling pathway will determine changes in H3K27ac. On the basis of the genome-wide H3K27ac changes we computed for each enhancer and each inhibitor a score representing the deviation from the vehicle alone. Cluster analysis of the Inhibitors allowed us to group the 58 inhibitors used into 7 clusters displaying coherent enrichments of transcription factor DNA recognition motifs. Based on this evidence and preliminary analysis of in vivo binding of kinase inhibitors to target kinases, we hypothesize that epigenomic effects of kinase inhibitors depend on a combination of on- and off-target effects.