Improvement of Peptide Separation for Exploring the Missing Proteins Localized on Membranes

Following an enormous effort by the
global scientific community
coordinated by HUPO’s Human Proteome Project, the number of
proteins without high-quality MS or other evidence (colloquially termed
missing proteins) has substantially decreased; however, some highly
hydrophobic MPs remain on the list. We believe that efficient peptide
separation is an approach that can be used to improve the identification
of these hydrophobic MPs. We propose that peptides prepared from the
membrane fractions of human cell lines and placental tissue can be
well separated from hydrophilic peptides in organic solvents at high
concentrations due to the precipitation of hydrophilic peptides with
lower solubility. Using a combination strategy of peptide separation
in 98% acetonitrile prior to traditional 2D reverse-phase liquid chromatography,
more hydrophobic peptides were detected in the supernatants of the
organic solvent extractions than were found in the pellets. When this
strategy was adopted, 30 MPs (≥2 non-nested unique peptides
with ≥9 amino acids) with 114 unique peptides were identified
at protein false discovery rate (FDR) < 1%, including 7, 12, and
13 MPs obtained from membrane preparations derived from K562, HeLa
cells, and human placenta, respectively. Of the 30 MPs identified
in this study, 19 were categorized as membrane proteins or extracellular
matrix proteins. Furthermore, 20 were verified to possess two non-nested
unique peptides through parallel reaction monitoring with the corresponding
chemically synthesized peptides. The use of organic solvents at high
concentrations was shown to be an efficient way to improve the exploration
of hydrophobic MPs. The data obtained in this study are available
via ProteomeXchange (PXD010630) and PeptideAtlas (PASS01218).