An atlas of human kinase regulation

Abstract

The coordinated regulation of protein kinases is a rapid mechanism that integrates diverse cues and swiftly determines appropriate cellular responses. However, our understanding of cellular decision‐making has been limited by the small number of simultaneously monitored phospho‐regulatory events. Here, we have estimated changes in activity in 215 human kinases in 399 conditions derived from a large compilation of phosphopeptide quantifications. This atlas identifies commonly regulated kinases as those that are central in the signaling network and defines the logic relationships between kinase pairs. Co‐regulation along the conditions predicts kinase–complex and kinase–substrate associations. Additionally, the kinase regulation profile acts as a molecular fingerprint to identify related and opposing signaling states. Using this atlas, we identified essential mediators of stem cell differentiation, modulators of Salmonella infection, and new targets of AKT1. This provides a global view of human phosphorylation‐based signaling and the necessary context to better understand kinase‐driven decision‐making.

The Tyrosine Kinase Inhibitor Gefitinib Restricts Mycobacterium tuberculosis Growth through Increased Lysosomal Biogenesis and Modulation of Cytokine Signaling

“…Approximately 60 phosphosites exhibited altered abundance in response to M. tuberculosis infection across the 4 and 8 h time points (Figure 2B). Comparison of changes in phosphotyrosine signaling resulting from M. tuberculosis infection with a curated database of published phosphoproteomics using PhosFate 26 strongly correlated with other phosphoproteomic studies of cells infected with the bacterial pathogens Shigella flexneri 27 and Salmonella enterica 28 in epithelial cells (Figure 2C). Interestingly, M. tuberculosis infection of macrophages also correlated with EGF treatment of epithelial cells and has an inverse correlation with epithelial cells activated with EGF and treated with a MAPK inhibitor 29 (Figure 2C).…”

“…Statistically significantly changing sites were selected by applying a log2-fold-change (>1.0) and an adjusted p value (<0.05) corrected for multiple testing threshold. Phosphotyrosine log2-fold-change profiles were uploaded to the PhosFate Profiler tool (Phosfate.com 26 ) to identify published phosphoproteomics data sets with significant correlations. Correlations with p values <0.05 are illustrated in Figure 2C.…”

Elucidating essential kinases of endothelin signalling by logic modelling of phosphoproteomics data

“…(iii) The network models resulting from the two analysed cell lines were strikingly similar despite the moderate overlap in phosphoproteomics data. These results are in line with a metaanalysis of publicly available phosphoproteomics data set, showing that the same cell perturbations in different studies led to the same kinase activation predictions, although the actual phosphopeptides underlying these predictions were different (Ochoa et al, 2016). The EDNRB signalling model is more expansive and coherent than the current heterogeneously assembled models of EDN signalling (Bouallegue et al, 2007;Rosano et al, 2013).…”

Systematic Analysis of Transcriptional and Post-transcriptional Regulation of Metabolism in Yeast

“…TF activities were inferred using a regulatory network obtained by combining gene-expression data from TF knock-out experiments and TF binding sites from ChIP-chip experiments (see Methods). The changes in activity of a regulator can be estimated by considering the changes of its targets [16,18,35]. For example, by analysing the phosphorylation changes of reported target sites of a protein K/P, one can predict whether the K/P is changing significantly (Fig 2A).…”