posted on 2012-07-06, 00:00authored byMathieu Courcelles, Gaëlle Bridon, Sébastien Lemieux, Pierre Thibault
The past decade has been marked by the emergence of selective
affinity
media and sensitive mass spectrometry instrumentation that facilitated
large-scale phosphoproteome analyses and expanded the repertoire of
protein phosphorylation. Despite these remarkable advances, the precise
location of the phosphorylation site still represents a sizable challenge
in view of the labile nature of the phosphoester bond and the presence
of neighboring phosphorylatable residues within the same peptide.
This difficulty is exacerbated by the combinatorial distribution of
phosphorylated residues giving rise to different phosphopeptide isomers.
These peptides have similar physicochemical properties, and their
separation by LC is often problematic. Few studies have described
the frequency and distribution of phosphoisomers in large-scale phosphoproteomics
experiments, and no convenient informatics tools currently exist to
facilitate their detection. To address this analytical challenge,
we developed two algorithms to detect separated and co-eluting phosphopeptide
isomers and target their subsequent identification using an inclusion
list in LC–MS/MS experiments. Using these algorithms, we determined
that the proportion of isomers present in phosphoproteomics studies
from mouse, rat, and fly cell extracts represents 3–6% of all
identified phosphopeptides. While conventional analysis can identify
chromatographically separated phosphopeptides, targeted LC–MS/MS
analyses using inclusion lists provided complementary identification
and expanded the number of phosphopeptide isomers by at least 52%.
Interestingly, these analyses revealed that the occurrence of phosphopeptides
isomers can also correlate with the presence of extended phosphorylatable
amino acids that can act as a “phosphorylation switch”
to bind complementary domains such as those present in SR proteins
and ribonucleoprotein complexes.