A Tandem Mass Spectrometry Sequence Database Search Method for Identification of O‑Fucosylated Proteins by Mass Spectrometry

Thrombospondin type
1 repeats (TSRs), small adhesive protein domains
with a wide range of functions, are usually modified with O-linked
fucose, which may be extended to O-fucose-β1,3-glucose.
Collision-induced dissociation (CID) spectra of O-fucosylated peptides
cannot be sequenced by standard tandem mass spectrometry (MS/MS) sequence
database search engines because O-linked glycans are highly labile
in the gas phase and are effectively absent from the CID peptide fragment
spectra, resulting in a large mass error. Electron transfer dissociation
(ETD) preserves O-linked glycans on peptide fragments, but only a
subset of tryptic peptides with low m/z can be reliably sequenced from ETD spectra compared to CID. Accordingly,
studies to date that have used MS to identify O-fucosylated TSRs have
required manual interpretation of CID mass spectra even when ETD was
also employed. In order to facilitate high-throughput, automatic identification
of O-fucosylated peptides from CID spectra, we re-engineered the MS/MS
sequence database search engine Comet and the MS data analysis suite
Trans-Proteomic Pipeline to enable automated sequencing of peptides
exhibiting the neutral losses characteristic of labile O-linked glycans.
We used our approach to reanalyze published proteomics data from Plasmodium parasites and identified multiple glycoforms
of TSR-containing proteins.