Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 May;16(5):786-798.
doi: 10.1074/mcp.M116.066233. Epub 2017 Mar 6.

Machine Learning of Global Phosphoproteomic Profiles Enables Discrimination of Direct versus Indirect Kinase Substrates

Evgeny Kanshin  1 , Sébastien Giguère  1 , Cheng Jing  1 , Mike Tyers  2   3 , Pierre Thibault  2   4
Affiliations

Machine Learning of Global Phosphoproteomic Profiles Enables Discrimination of Direct versus Indirect Kinase Substrates

Evgeny Kanshin et al. Mol Cell Proteomics. 2017 May.

Abstract

Mass spectrometry allows quantification of tens of thousands of phosphorylation sites from minute amounts of cellular material. Despite this wealth of information, our understanding of phosphorylation-based signaling is limited, in part because it is not possible to deconvolute substrate phosphorylation that is directly mediated by a particular kinase versus phosphorylation that is mediated by downstream kinases. Here, we describe a framework for assignment of direct in vivo kinase substrates using a combination of selective chemical inhibition, quantitative phosphoproteomics, and machine learning techniques. Our workflow allows classification of phosphorylation events following inhibition of an analog-sensitive kinase into kinase-independent effects of the inhibitor, direct effects on cognate substrates, and indirect effects mediated by downstream kinases or phosphatases. We applied this method to identify many direct targets of Cdc28 and Snf1 kinases in the budding yeast Saccharomyces cerevisiae Global phosphoproteome analysis of acute time-series demonstrated that dephosphorylation of direct kinase substrates occurs more rapidly compared with indirect substrates, both after inhibitor treatment and under a physiological nutrient shift in wt cells. Mutagenesis experiments revealed a high proportion of functionally relevant phosphorylation sites on Snf1 targets. For example, Snf1 itself was inhibited through autophosphorylation on Ser391 and new phosphosites were discovered that modulate the activity of the Reg1 regulatory subunit of the Glc7 phosphatase and the Gal83 β-subunit of SNF1 complex. This methodology applies to any kinase for which a functional analog sensitive version can be constructed to facilitate the dissection of the global phosphorylation network.

PubMed Disclaimer

Figures

Fig. 1.
Fig. 1.
Global effects of Cdc28-as inhibition. (a) Addition of inhibitor to cells bearing as-form of a kinase can result in three types of phosphorylation events represented by colored arrows. The inhibitor itself can affect cellular homeostasis leading to changes in the phosphoproteome (kinase-independent). Upon inhibition of the target kinase its direct substrates will be dephosphorylated, which in turn can affect phosphorylation of other downstream targets, termed indirect substrates). (b) Yeast cells containing the cdc28-as allele were treated with vehicle or 10 μm 1NM-PP1 for 15 min. Experiments were performed in six replicates and only phosphopeptides quantified by MS in at least four replicates were considered for subsequent analysis. Significant (FDR<5%) changes were observed for 1,728 phosphosites without a clear prevalence of dephosphorylation (1,008) over phosphorylation (720) but with enrichment of CDK consensus motif among dephosphorylated phosphosites.
Fig. 2.
Fig. 2.
1NM-PP1 causes off-target phosphorylation events. (a) In order to estimate the effect of 1NM-PP1 treatment alone a wild type (non-as) strain was treated with 10 μm 1NM-PP1 for 15 min and compared with DMSO control treatment. Significant changes were observed on ∼13% of detected phosphosites. (b) MotifX analysis revealed basophilic kinase motifs among the dephosphorylated sites in a 1NM-PP1 treated wt strain. NetworKIN analysis also suggested that ∼60% of dephosphorylated sites were predicted substrates of the basophilic kinase PKA (Tpk1, Tpk2, and Tpk3).
Fig. 3.
Fig. 3.
Cdc28-specific phosphorylation sites. (a) Subtraction of phosphosites due to 1NM-PP1 alone from a cdc28-as inhibition experiment revealed targets whose abundance was affected by the change in Cdc28 activity. In total ∼12% of all identified phosphosites were regulated by Cdc28 with strong preference for dephosphorylation (523 dephosphorylated versus 207 phosphorylated sites). The increase in phosphorylation on 207 substrates indicates existence of indirect downstream effectors of Cdc28 inhibition. (b) Removal of 1NM-PP1 off-target phosphorylation events (1NM-PP1) from the Cdc28as results (global) improved the quality of the dataset as indicated by dramatic increase of the ratio of known Cdc28 substrates and sites with a CDK consensus motif between down-regulated and up-regulated peptides (Cdc28). (c) NetworKIN analysis revealed enrichment of CDK substrates among dephosphorylated phosphosites.
Fig. 4.
Fig. 4.
A machine learning algorithm separates direct from indirect Cdc28 targets. (a) Correction for the 1NM-PP1 off-target effect yielded a list of dephosphorylated sites that correspond to both direct and indirect Cdc28 targets. A mixture of direct and indirect Cdc28 targets (dephosphorylated phosphosites) along with part of the negative control set of up-regulated phosphosites across all experiments was used to build a support vector machine (SVM) model. Accuracy was estimated by cross-validation on the remaining part of the negative control group. Known CDK substrates as well as CDK motifs were enriched among direct targets predicted by the model. (b) The KID was used to extract known Cdc28 partners from literature-based studies, which were compared with direct and indirect Cdc28 sites predicted by the SVM model.
Fig. 5.
Fig. 5.
Predicted direct Cdc28 substrates are dephosphorylated more rapidly than predicted indirect substrates. (a) Correlation between extent of site dephosphorylation and SVM score. (b) Extent of dephosphorylation was larger for direct versus indirect substrates (pValue = 1e-8, Wilcoxon test). (c) Kinetic measurements of phosphorylation changes upon Cdc28 inhibition over a 15-min time course at 1 min intervals revealed differential dephosphorylation dynamics between direct versus indirect Cdc28 targets.
Fig. 6.
Fig. 6.
Identification of direct Snf1 substrates. (a) Application of global phosphorylation profiles and machine learning to identify direct targets of the Snf1 kinase. After removal of 1NM-PP1 inhibitor off-target phosphorylation events, significant regulation was observed for ∼ 5% of all detected sites with a strong prevalence of dephosphorylation. SVM model allowed separation of direct (134) and indirect (78) Snf1 targets among all dephosphorylated targets. (b) MotifX analysis revealed enrichment of a sequence motif where the phosphorylated residue was followed by leucine in the +4 position for direct Snf1 substrates. No significantly enriched motifs were obtained in a similar analysis of indirect Snf1 targets. Gene ontology analysis revealed enrichment of terms associated with sugar metabolism and regulation of cell growth. (c) Many known components of the local Snf1 signaling network were identified as either direct or indirect targets.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Pawson T., and Scott J. D. (2005) Protein phosphorylation in signaling—50 years and counting. Trends Biochem. Sci. 30, 286–290 - PubMed
    1. Hunter T. (2000) Signaling—2000 and beyond. Cell 100, 113–127 - PubMed
    1. Hunter T., and Plowman G. D. (1997) The protein kinases of budding yeast: Six score and more. Trends Biochem. Sci. 22, 18–22 - PubMed
    1. Macek B., Mann M., and Olsen J. V. (2009) Global and site-specific quantitative phosphoproteomics: Principles and applications. Annu. Rev. Pharmacol. Toxicol. 49, 199–221 - PubMed
    1. Rigbolt K. T., Prokhorova T. A., Akimov V., Henningsen J., Johansen P. T., Kratchmarova I., Kassem M., Mann M., Olsen J. V., and Blagoev B. (2011) System-wide temporal characterization of the proteome and phosphoproteome of human embryonic stem cell differentiation. Sci. Signal. 4, rs3. - PubMed

MeSH terms

Substances

LinkOut - more resources