Quantitative Phospho-Proteomics in Early Stem Cell Differentiation

Javier Muñoz *, Dennis Van Hoof , Stefan R. Braam , Martijn W.H. Pinkse , Rune Linding , Albert J.R. Heck , Christine L. Mummery ,


Pluripotent embryonic stem cells (ESC) can self-renew indefinitely and possess characteristic protein-protein networks which remodel during differentiation. How this is controlled at the protein and protein-signalling level is poorly understood. To study the networks and cascades activated by phosporylation in the early phase of differentiation, we adopted a quantitative proteomic approach in ESC, combining SILAC labeling and selective capturing of phospho-peptides. Embryonic stem cells were SILAC-labeled with Lys and Arg. Unstimulated labeled cells were mixed with unlabeled cells that had been incubated with bone morphogenetic protein (BMP) for 30, 60 or 240 minutes to induce differentiation. From each of these combined samples, proteins were digested and phosphopeptides were enriched by subsequent SCX and on-line TiO2 chromatography. Proteins were identified using Mascot, and quantified using MSQuant. In total, 144 LC-MS runs were performed using LTQ-FT and LTQ-orbitrap mass spectrometry, collecting over 800,000 peptide spectra. Over 2000 phosphopeptides were identified in each of the 3 datasets, totaling over 3600 unique phosphopeptides across all samples. Next, phosphorylation levels were quantified across the 3 time points, and data were analysed at various levels. First, we have followed the dynamics of phosphorylation over time at the peptide level, since for most peptides we have mapped the phosphorylation site. Indeed we have found many different profiles, which often varied foro individual phosphorylation sites in the same protein. Second, we have deduced kinase motifs enriched in our dataset, based on conserved regions around phosphorylation sites using NetworKIN. This identified CDK1/2 as a central kinase in the activated network. Third, we have mapped (differential) phosphorylation to signaling pathways by using gene ontologies and prior knowledge of protein-protein and biochemical interactions. This has indicated the activation of several signaling cascades, including Jnk and Akt. Finally, we have identified 26 proteins that are widely used as stem cell markers because they are associated with the undifferentiated state of ESC. In 12 of these proteins, 30 phosphorylation sites were identified, of which 25 were unknown so far. This could indicate that (de)phosphorylation of these proteins could be associated with the induction of differentiation. Currently we are integrating these combined data in a model that describes the events in the initial phase of ESC differentiation.