A Methodological Assessment and Characterization of Genetically-Driven Variation in Three Human Phosphoproteomes

Abstract

Phosphorylation of proteins on serine, threonine, and tyrosine residues is a ubiquitous post-translational modification that plays a key part of essentially every cell signaling process. It is reasonable to assume that inter-individual variation in protein phosphorylation may underlie phenotypic differences, as has been observed for practically any other molecular regulatory phenotype. However, we do not know much about the extent of inter-individual variation in phosphorylation because it is quite challenging to perform a quantitative high throughput study to assess inter-individual variation in any post-translational modification. To test our ability to address this challenge with SILAC-based mass spectrometry, we quantified phosphorylation levels for three genotyped human cell lines within a nested experimental framework, and found that genetic background is the primary determinant of phosphoproteome variation. We uncovered multiple functional, biophysical, and genetic associations with germline driven phosphopeptide variation. Variants affecting protein levels or structure were among these associations, with the latter presenting, on average, a stronger effect. Interestingly, we found evidence that is consistent with a phosphopeptide variability buffering effect endowed from properties enriched within longer proteins. Because the small sample size in this 'pilot' study may limit the applicability of our genetic observations, we also undertook a thorough technical assessment of our experimental workflow to aid further efforts. Taken together, these results provide the foundation for future work to characterize inter-individual variation in post-translational modification levels and reveal novel insights into the nature of inter-individual variation in phosphorylation.

Figure 1

Nested SILAC-based phosphoproteomic analysis. (A) HapMap LCLs derived from three Yoruba males in Ibadan, Nigeria were repeatedly cultured and repeatedly subjected to a multistep mass spectrometry workflow. Dotted vs un-dotted circles represent the two different processing batches. (B) Protein extract from each stable isotope labeled sample is paired with an equal amount of extract from a common unlabeled cell line. This mixture is digested, separated into fractions via strong cation exchange (SCX) chromatography, and phosphopeptides are enriched with TiO₂ resin and subjected to 1.5 hour HPLC runs on a Q-Exactive hybrid quadrupole-orbitrap mass spectrometer.

Figure 2

Genetically-driven phosphoproteomic variation. Violin plots of (A) absolute and (B) standardized phosphopeptide variance components derived from each layer of the hierarchical design after accounting for protein levels. (C) Heatmap of protein normalized phosphopeptide SILAC ratios. (D) Venn diagram of differential phosphorylation results across all three pairwise inter-individual comparisons (FDR 5%).