Conserved phosphorylation hotspots in eukaryotic protein domain families

Abstract

Protein phosphorylation is the best characterized post-translational modification that regulates almost all cellular processes through diverse mechanisms such as changing protein conformations, interactions, and localization. While the inventory for phosphorylation sites across different species has rapidly expanded, their functional role remains poorly investigated. Here, we combine 537,321 phosphosites from 40 eukaryotic species to identify highly conserved phosphorylation hotspot regions within domain families. Mapping these regions onto structural data reveals that they are often found at interfaces, near catalytic residues and tend to harbor functionally important phosphosites. Notably, functional studies of a phospho-deficient mutant in the C-terminal hotspot region within the ribosomal S11 domain in the yeast ribosomal protein uS11 shows impaired growth and defective cytoplasmic 20S pre-rRNA processing at 16 °C and 20 °C. Altogether, our study identifies phosphorylation hotspots for 162 protein domains suggestive of an ancient role for the control of diverse eukaryotic domain families.

SAPH-ire TFx – A Recommendation-based Machine Learning Model Captures a Broad Feature Landscape Underlying Functional Post-Translational Modifications

“…Previous models (including our own) have employed the use of quantitative features in isolation (15,16) or in combination (20,22–24,36)) to rank PTMs according to potential for function. In the process of re-evaluating this approach, we concluded that rank-based models are incongruous with functional evidence from biological data, wherein functional impact is broadly defined and highly variable.…”

“…Indeed, this underlying dilemma tends to promote a perspective that the study of PTMs is risky and quite possibly not worth the effort, especially in light of the exponential increase in data content that can appear overwhelming to a protein experimentalist. PTM site features such as evolutionary conservation (14–16), PTM co-localization (17–19), protein structural constraints (20,21), and other single features, have been shown to be mildly predictive for function and offer useful means of filtering PTM data for likely important modifications. More recently, machine learning models that incorporate multiple single features have been proven to have significantly greater predictive power in comparison to single features alone (22–24).…”

Proteome-wide inference of protein kinase regulatory circuits

“…To do so we first evaluated how phosphorylation at a specific site is likely to affect a given kinase's activity. We found particular discrimination for sites within phosphorylation hotspots of the protein-kinase domain (Strumillo et al, 2019): sites within hotspots tend overwhelmingly to be activating (i.e. within the kinase activation loop) ( Figure 3a, 1st panel).…”

“…As features, we used: the percentage position of the site relative to the start and end of the protein kinase domain (i.e. between 0 and 1 for sites that fall within the domain); the percentage position of the site along the protein's length; the domain (if any) in which the phosphosite lies, including, but not limited to, protein kinase domains; the phosphosite residue (serine/threonine or tyrosine); whether or not the substrate is a tyrosine kinase; an estimate of secondary sequence disorder, as calculated by DISOPRED (Ward et al, 2004); and the −log 10 p-value of the site being in a phosphorylation hot-spot (Strumillo et al, 2019).…”

Towards a systematic map of the functional role of protein phosphorylation

“…For example, prediction of phosphosites that may regulate protein interactions or cross-regulatory interactions with other PTMs have been tried (Nishi et al , 2011;Beltrao et al , 2012;Šoštarić et al , 2018) . In addition, the conservation of phosphosites across species has been used to infer functionality, as phosphosites with well characterized functions are more likely to be conserved (Studer et al , 2016;Strumillo et al , 2019) . In contrast, experimental approaches that functionally characterize PTM sites in a comprehensive manner are still lacking or are geared to a small number of phosphosites (Nakic et al , 2016;Oliveira et al , 2015) .…”