Large-Scale Mass Spectrometric Detection of Variant Peptides Resulting from Nonsynonymous Nucleotide Differences

Each
individual carries thousands of nonsynonymous single nucleotide
variants (nsSNVs) in their genome, each corresponding to a single
amino acid polymorphism (SAP) in the encoded proteins. It is important
to be able to directly detect and quantify these variations at the
protein level to study post-transcriptional regulation, differential
allelic expression, and other important biological processes. However,
such variant peptides are not generally detected in standard proteomic
analyses due to their absence from the generic databases that are
employed for mass spectrometry searching. Here we extend previous
work that demonstrated the use of customized SAP databases constructed
from sample-matched RNA-Seq data. We collected deep-coverage RNA-Seq
data from the Jurkat cell line, compiled the set of nsSNVs that are
expressed, used this information to construct a customized SAP database,
and searched it against deep-coverage shotgun MS data obtained from
the same sample. This approach enabled the detection of 421 SAP peptides
mapping to 395 nsSNVs. We compared these peptides to peptides identified
from a large generic search database containing all known nsSNVs (dbSNP)
and found that more than 70% of the SAP peptides from this dbSNP-derived
search were not supported by the RNA-Seq data and thus are likely
false positives. Next, we increased the SAP coverage from the RNA-Seq
derived database by utilizing multiple protease digestions, thereby
increasing variant detection to 695 SAP peptides mapping to 504 nsSNV
sites. These detected SAP peptides corresponded to moderate to high
abundance transcripts (30+ transcripts per million, TPM). The SAP
peptides included 192 allelic pairs; the relative expression levels
of the two alleles were evaluated for 51 of those pairs and were found
to be comparable in all cases.