Biology Reference
In-Depth Information
baits), as well as many unique interactors. Because the
baits were purified under two conditions, baseline and
insulin stimulation, the dynamics of the mini-proteome
during signaling events was uncovered. As a measure of
the sensitivity of the TAP/MS approach to network char-
acterization, interactions among the canonical compo-
nents and their known interactors were extracted. This
canonical network recapitulated most of the known RTK-
ERK signaling pathway.
Comparing the RTK-ERK PPI network to six unbiased
genome-wide RNAi screens revealed that nearly half (119)
of the proteins identified by PPI mapping scored in the
RNAi screens, which is a significant enrichment relative to
the entire genome (19%, p
attachment of a unique tag or label, which enables the
quantification of the same peptide across multiple samples.
The uniquely labeled samples are combined and run
through an MS analysis. Despite bearing distinct tags, the
same peptides from the different samples are indistin-
guishable from each other in the first MS run, because the
molecular weight of each tag is the same. However, during
MS/MS each tag undergoes fragmentation, releasing
a signature reporter ion. The signature reporter ions differ
in mass between the tags and their relative levels serve as
a measure of differences in the levels of a given peptide
between samples. Labeling methods such as these provide
a rapid means by which to quantitatively examine global
proteome-level changes, and compare, for example, wild-
type cells with those subjected to mutations, RNAi
knockdown or small molecule treatments.
10
e
25
)
[187]
.
A major bottleneck in large-scale proteomics studies is
the experimental validation of specific interactors or
components of complexes. The combination of AP-MS
with RNAi-mediated knockdown provides a way to
directly validate specific PPIs. With differential labeling of
the two proteomes to be compared (wild-type vs. RNAi
knockdown) such analyses have the potential for accurate,
highly quantitative results
[201,202]
. Currently, two major
types of labeling technique are used for MS-based pro-
teomics studies: metabolic labeling and chemical labeling
[203]
. Stable isotope labeling by amino acids in cell
culture (SILAC) is considered to be the gold standard in
the case of metabolic labeling. Here, isotopically labeled
amino acids (for example arginine and lysine labeled with
the stable
13
C and/or
15
N isotope) are incorporated into
cellular proteins during normal protein biosynthesis
[204]
.
Thus, the cells to be compared (for example wild-type vs.
RNAi) are grown in media containing 'light' (normal) and
'heavy' (labeled) amino acids, respectively. After labeling,
the two cell populations are mixed, fractionated, and
subject to MS/MS analysis to quantify the differences
between their two proteomes in a highly accurate manner
[205,206]
. Because the labels are carried by arginine and
lysine residues tryptic digestion produces peptides that
contain a labeled amino acid at the carboxy terminus. The
heavy and light tryptic peptides elute together as pairs
separated by a defined mass difference that allows the two
proteomes to be distinguished in the MS/MS analyses.
SILAC can be multiplexed to allow comparisons between
three different proteomes simultaneously. SILAC-based
differential labeling combined with RNAi, co-immuno-
precipitation and quantitative MS analysis was used to
detect and validate the cellular interaction partners of
endogenous
b
-catenin and Cbl proteins in mammalian
cells
[202]
. Alternatively, chemical labeling involves the
use of isobaric tagging reagents such as iTRAQ (isobaric
tags for relative and absolute quantification)
[207]
or TMT
(tandem mass tags)
[208]
(see
Box 5.1
for definitions) to
label peptides after lysis and trypsinization. The peptides
in samples to be compared are modified by covalent
<
7
Transcriptional Profiling
Gene expression profiling using DNA microarrays, and
more recently RNA-seq, has emerged as a valuable tool for
broad correlation of gene activity with alterations in
physiological or developmental states
[209
211]
.Tran-
scriptional profiling experiments can be used to generate
compendia of gene expression data across different cell
types
[212]
, diverse species
[213]
, development times
[214]
, and in response to distinct stimuli
[215]
. Such gene
expression datasets have been commonly used to identify
genes that function in common pathways or which encode
components of the same complex. Studies in yeast have
demonstrated that proteins that interact with each other
show similar
e
expression profiles
to non-interacting
proteins
[216
219]
. Gene expression profiling has been
used to study signaling by wild-type and mutant receptor
tyrosine kinases (RTKs) and has provided evidence for
substantially overlapping immediate early transcriptional
responses upon activation of PLC
g
, PI3K, SHP2, and
RasGAP proteins and their respective signaling pathways
[220]
. However, expression profiling studies do not
provide details of how and where in this network pathways
engage in cross-talk to specify the appropriate biological
response.
Although expression profiling studies have become the
gold standard for global responses to signaling, several
recent studies have shown that correlation between tran-
scriptome and proteome is only ~50%
[221,222]
. Proteo-
mics approaches that directly measure the targets of
signaling pathways
e
are more
useful. For example, Yates and colleagues
[223]
compared
protein abundance between wild-type C. elegans and those
lacking the worm insulin receptor (InR) ortholog daf-2.
This study revealed 86 proteins whose abundance changed
following loss of InR, an important finding for a signaling
that is, the proteins
e
e
