TY - JOUR
T1 - Predicting Kinase Activity in Angiotensin Receptor Phosphoproteomes Based on Sequence-Motifs and Interactions
AU - Bøgebo, Rikke
AU - Horn, Heiko
AU - Olsen, Jesper V
AU - Gammeltoft, Steen
AU - Jensen, Lars J
AU - Hansen, Jakob L
AU - Christensen, Gitte Lund
PY - 2014/4/10
Y1 - 2014/4/10
N2 - Recent progress in the understanding of seven-transmembrane receptor (7TMR) signalling has promoted the development of a new generation of pathway selective ligands. The angiotensin II type I receptor (AT1aR) is one of the most studied 7TMRs with respect to selective activation of the β-arrestin dependent signalling. Two complimentary global phosphoproteomics studies have analyzed the complex signalling induced by the AT1aR. Here we integrate the data sets from these studies and perform a joint analysis using a novel method for prediction of differential kinase activity from phosphoproteomics data. The method builds upon NetworKIN, which applies sophisticated linear motif analysis in combination with contextual network modelling to predict kinase-substrate associations with high accuracy and sensitivity. These predictions form the basis for subsequently nonparametric statistical analysis to identify likely activated kinases. This suggested that AT1aR-dependent signalling activates 48 of the 285 kinases detected in HEK293 cells. Of these, Aurora B, CLK3 and PKG1 have not previously been described in the pathway whereas others, such as PKA, PKB and PKC, are well known. In summary, we have developed a new method for kinase-centric analysis of phosphoproteomes to pinpoint differential kinase activity in large-scale data sets.
AB - Recent progress in the understanding of seven-transmembrane receptor (7TMR) signalling has promoted the development of a new generation of pathway selective ligands. The angiotensin II type I receptor (AT1aR) is one of the most studied 7TMRs with respect to selective activation of the β-arrestin dependent signalling. Two complimentary global phosphoproteomics studies have analyzed the complex signalling induced by the AT1aR. Here we integrate the data sets from these studies and perform a joint analysis using a novel method for prediction of differential kinase activity from phosphoproteomics data. The method builds upon NetworKIN, which applies sophisticated linear motif analysis in combination with contextual network modelling to predict kinase-substrate associations with high accuracy and sensitivity. These predictions form the basis for subsequently nonparametric statistical analysis to identify likely activated kinases. This suggested that AT1aR-dependent signalling activates 48 of the 285 kinases detected in HEK293 cells. Of these, Aurora B, CLK3 and PKG1 have not previously been described in the pathway whereas others, such as PKA, PKB and PKC, are well known. In summary, we have developed a new method for kinase-centric analysis of phosphoproteomes to pinpoint differential kinase activity in large-scale data sets.
U2 - 10.1371/journal.pone.0094672
DO - 10.1371/journal.pone.0094672
M3 - Journal article
C2 - 24722691
SN - 1932-6203
VL - 9
SP - 1
EP - 9
JO - PLoS Computational Biology
JF - PLoS Computational Biology
IS - 4
M1 - e94672
ER -