Biology Reference
In-Depth Information
feedforward loops are likely to be involved in information
processing, while interactions that are part of bi-parallel
motifs are likely to be involved in energy transfer. We
anticipate that analytical tools utilizing network motifs,
particularly multicolor network motifs, will help in
deciphering the function of many interactome network
edges and local structures.
Multicolor triangle motifs containing two nodes linked
by protein interactions with both of these nodes connected
to a third node by a genetic interaction appear enriched in
the S. cerevisiae interactome
[24]
. Consequently, if A, B
and C are three genes such that the translation products of A
and B physically interact, and the A and C genes are linked
by a genetic interaction, then a genetic interaction between
B and C can be predicted
[131]
. This particular motif
suggests a 'compensatory complex' theme wherein two
proteins/complexes/processes function in parallel. An
excellent example of compensatory complexes is the pair-
ing of endoplasmic reticulum (ER) protein-translocation
sub-complex
[190]
and the Gim complex
[191]
, with each
complex connected to the other by multiple genetic inter-
actions
[24,150]
. The Gim complex facilitates the folding
of actin and tubulin components of the cytoskeleton. The
genetic interactions between the Gim complex and ER
protein translocation suggests that defects in moving
proteins into the ER are ameliorated by a fully functioning
cytoskeleton, whereas the trafficking of protein via lipid
vesicles requires the cytoskeleton to act as a 'molecular
train track'.
and quantity by chaperones, and actively targeted to the
required site of action. There, protein complex assembly
can require a specific order of addition to reach stability.
Because each step leading to a protein complex is poten-
tially subject to regulation, co-complex interactome
networks are dense with accumulated information on the
cell dynamics.
Interrogating the dynamics of complex assembly at
proteome scale is not yet feasible experimentally. It is,
however, becoming possible to compare proteome expres-
sion across cell types, thanks to technological innovations
developed throughout the last decade, such as stable
isotope labeling by amino acids in cell culture
[192]
(see
Chapter 1). We can now interrogate interactome networks
for node dynamics (at least partially), but not yet for edge
dynamics. The first dynamic measures of protein complex
membership successfully followed a single protein, GRB2,
as it dynamically associated with multiple complexes
[193]
. Pending increases in throughput and further
advances, the modeling of co-complex interactome
network dynamics will need to rely on computational
analyses. Empirical measurements of binary interaction
dynamics are also lacking. The LUMIER technology has
paved the way
[194]
, but most binary interactome maps
remain static, and attempts at dynamical modeling also rely
on computational analyses.
Computational integration of interactome maps and
expression profiles can identify biological conditions
whereby two proteins that can interact, according to an
interactome map, are also co-present, according to their
expression profiles. This additional knowledge allows the
inference of spatiotemporal 'interaction territories'
marking where or when the interaction can take place,
e.g., during cell cycle or organism development
[52,122]
.
To what extent is the expression of interacting proteins
transcriptionally coordinated in cellular systems? Physi-
cally interacting proteins are more likely to exhibit similar
expression patterns than would be expected by chance
[39,129,195]
. Most interacting proteins are not co-
expressed, however, and some pairs are even anti-
correlated in expression. Interactome dynamics therefore
appear to be under tight transcriptional control, with most
protein interactions being transient.
Transient protein interactions involved in signaling
and intercomplex connections are enriched in binary
interactome maps compared to co-complex interactome
maps
[39]
. Members of a given protein complex can be
co-regulated by a common transcription factor, when
a transcription factor is connected by transcription regu-
latory edges directed towards several interacting and
co-expressed proteins, forming 'regulonic complexes'
[24,196,197]
. In response to extracellular perturbations,
protein complexes generally remain stable, but the
functional connections between these complexes are
TOWARDS DYNAMIC INTERACTOMES
Towards Cell-Type and Condition-Specific
Interactomes
The cell interior is a constantly changing environment.
Biomolecules and cellular processes respond dynamically
to intra- and extracellular cues. Available protein
protein
interactome maps are, regrettably, mostly static, repre-
senting the union of protein interactions that may occur in
all locations, times and environments. Analysis of pro-
tein
e
protein interactome networks will continue to
contribute profound insights to systems biology only by
reaching the temporal and spatial resolution necessary to
dynamically model coordination of biological processes
across the cell and the organism.
For a protein complex to be active at the right time and
place in the cell, and at a controlled concentration, the cell
has to undertake a large number of parallel and successive
decisions. For each complex subunit, the cognate gene
needs to be transcribed (chromatin opening, initiation and
elongation of transcription
e
) and the mRNA processed,
exported, and translated. Complex subunits often also need
to be post-translationally modified, controlled in quality
.
