Back End Strategies To PIK-3

This strategy builds upon existing methods based on identification of kinase consensus motifs (Linding et?al., 2008; Obata et?al., 2000; Obenauer et?al., 2003) or chemical-genetic approaches (Blethrow et?al., 2008; Chi et?al., 2008) while overcoming problems associated with consensus motif redundancy and preserving the specificity and context of the cellular environment. To characterize the insulin signaling network we applied two distinct quantitative phosphoproteomics strategies using stable isotope labeling with amino acids in cell culture (SILAC) (Ong and Mann, 2006). In the first set of experiments, we quantified the insulin-regulated phosphoproteome using two potent and selective kinase inhibitors targeting Akt (MK2206) and PI3K/mTOR (LY294002) (Figure?1A). In the second series of experiments we employed a multiplexed SILAC approach to interrogate the temporal profile of insulin signaling over a wide temporal dynamic range spanning PIK-3 15?s to 60?min (Figure?1B). All MS experiments were performed in triplicate. We achieved high confidence and deep coverage of the phosphoproteome by analyzing a large number of phosphopeptide-enriched fractions and by taking advantage of recent developments in MS hardware (Michalski et?al., 2011) and software (Cox and Mann, 2008; Cox et?al., 2011) (Figure?1C). In total, 7,441,138 high-resolution (Orbitrap higher-energy collision dissociation [HCD]) spectra were acquired, resulting in the identification of 38,901 unique phosphopeptides corresponding to 37,248 phosphorylation selleck chemical sites on 5,705 proteins (Figure?1D and Tables S1 and S2), placing this among the largest phosphoproteomes reported. Phosphopeptide identification confidence was high, with 95% of all peptides identified having an absolute mass deviation less than 2 ppm (Figures S1A and S1B). Our phosphopeptide enrichment was highly efficient and selective, since phosphopeptides consisted of an average of 83% of the total peptides identified from the phosphopeptide-enriched samples (Figure?1E) and reproducibility of phosphopeptide quantitation between biological replicates was good (Figure?1F). Single-, double-, triple-, and higher-phosphorylated peptides represented 72%, 23%, 4% and phosphatase inhibitor library site localization probability score (Olsen et?al., 2006). We divided the phosphoproteome into four categories based on this probability score: class I (>0.75), class II (0.75�C0.5), class?III (0.5�C0.25), and class IV (