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

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 phosphorylation that is mediated by downstream kinases. Here, we describe a framework for assignment of direct 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 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 cells. Mutagenesis experiments revealed a high proportion of functionally relevant phosphorylation sites on Snf1 targets. For example, Snf1 itself was inhibited through autophosphorylation on Ser 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.

Synthesizing Signaling Pathways from Temporal Phosphoproteomic Data

“…One YHR131C phosphosite is dependent on Rck2 during osmotic stress (Romanov et al, 2017), supporting our Rck2 -> YHR131C prediction. Similarly, we verified that 67 out of 91 (74%) predicted Cdc28 targets have at least one phosphosite with defective phosphorylation following Cdc28 inhibition (Holt et al, 2009;Kanshin et al, 2017) (Supplemental Results).…”

Delineating the contribution of Spc105-bound PP1 to spindle checkpoint silencing and kinetochore microtubule attachment regulation

“…The budding yeast Spc105 contains two basic patches: an anterior one spanning residues 101-104 and a posterior one spanning residues 340-343. Notably, the anterior conserved, four-residue basic patch with the sequence RRRK may be subject to phosphoregulation, because the serine and threonine residues located immediately downstream from the basic patch are phosphorylated by mitotic and S-phase kinases (pSYT repository of phosphorylated peptides hosted at the Global Proteome Machine Database; http:// gpmdb.thegpm.org/psyt/index.html; original observations from Kanshin et al [2017] and Smolka et al [2007]). Therefore, to understand the activity of the basic patch, we created the basic patch mutant (BPM) of Spc105, or Spc105 BPM , wherein the basic residues in both basic patches are replaced with nonpolar alanine residues (101-RRRK-104::AAAA, 340-KRRK-343::AAAA).…”

Minimization of a harmful cross-talk between mitotic checkpoint silencing and error correction

“…Spc105 contains two basic patches, an anterior one spanning residues 101-104 and a posterior one spanning residues 340-343. Notably, the anterior conserved patch of four basic residues 'RRRK' may be subject to phosphoregulation, because the Serine and Threonine residues located immediately downstream from the basic patch are phosphorylated by mitotic and S-phase kinases [pSYT repository of phosphorylated peptides hosted at the Global Proteome Machine Database; original observations from (Kanshin et al, 2017;Smolka et al, 2007)]. Therefore, to understand the activity of the basic patch, we created the basic patch mutant (BPM) of Spc105, Spc105 BPM , wherein the basic residues in both basic patches are replaced with non-polar Alanine residues.…”