Biology Reference
In-Depth Information
protein interactions based on sequence similarity measures
across species
[217]
.
Single amino acid changes can result
pro-
tein interactome networks with the existence of universal
processes found within all forms of Life? Beyond the union
of individual interactions, interactome networks exhibit
higher-level organizational properties, such as signaling
pathways, or other types of functional module. Several
pathways and modules appear evolutionarily conserved, as
measured by orthology-based network alignment algorithms
[234,235]
. Similarly, topology-based network alignment
algorithms have revealed considerable similarities in the
local wiring of cellular networks across evolutionarily
distant organisms
[236]
. The global network topology of
binary interactomes of organisms as diverse as humans,
plants, worms and yeasts appear qualitatively similar, char-
acterized by a scale-free distribution of degrees and small-
world structures
[55]
(see Chapter 9). Likewise, the esti-
mated ratio of interacting pairs among all possible protein
pairs in these organisms, with genomes encoding anywhere
from 6000 to 30 000 proteins, appears surprisingly stable,
with 5
How to reconcile the dynamic rewiring of protein
e
in edgetic
perturbations of protein
protein interactome networks
[184]
. Fixation of such sequence changes under selective
constraints is expected to shape protein interaction inter-
faces. In agreement, the sequences of hub proteins appear
under tighter constraints than the sequences of non-hub
proteins, with intra-module hubs significantly more con-
strained than inter-module hubs
[202,218,219]
. The yeast
protein Pbs2, which endogenously interacts specifically
with a single yeast SH3 domain, is able to promiscuously
interact with many non-yeast SH3 domains
[220]
.Atan
equivalent level of sequence similarity, protein interactions
are more conserved within species, when considering
paralogous protein pairs originating from duplication
events, than across species when considering orthologous
protein pairs
[217]
. It seems that tinkering with interaction
interfaces and specificity causes protein interactions to co-
evolve dynamically within biological systems.
Immediately following gene duplication events, paralo-
gous proteins are expected to have identical protein
sequences and to share all of their interactors. Empirical
observations have revealed that the fraction of interactors
shared by paralogous proteins decreases over evolutionary
time, likely reflecting the well-described functional diver-
gence of retained paralogous proteins
[221]
. Such evolu-
tionary dynamics may explain the origin of the scale-free
degree
e
10 interactions per 10 000 protein pairs
[55]
.Itis
possible that these high-level systems properties are ulti-
mately the object of evolutionary conservation and so unify
all forms of Life.
In summary, natural selection seems to shape the
dynamic
e
protein interactome
networks. Regulatory interactome networks seem to evolve
faster
evolution
of
protein
e
protein interactome networks
[213,228]
. More refined models of the evolution of bio-
logical systems, including population size effects and the
concept of 'genotype networks', are being investigated
[237]
(see chapter by A. Wagner). Life could be perceived
as a system containing genotypes and phenotypes, with
genotypes shaping phenotypes through the prism of inter-
actomes, and phenotypes shaping genotypes through the
feedback of evolution by natural selection.
than
protein
e
protein interactome
networks invariably follow, via an evolutionary version of
the 'rich-get-richer' principle
[222,223]
(see Chapter 9).
These evolutionary dynamics may also lead to an elevated
clustering in protein
distribution
that
protein
e
protein interactome networks if self-
interactions are taken into account, as their duplication
enables the formation of novel complexes of paralogous
proteins
[224,225]
(see Chapter 9). The proteasome complex
likely evolved from an ancestral homodimeric interaction
through multiple successive duplication events
[25,226]
.
Attempts to estimate the interactome rewiring rate
following duplication events have yielded conflicting
results
[212,213,221
e
CONCLUDING REMARKS
It
protein
interactome mapping and modeling will be key to under-
standing
is becoming increasingly clear that protein
e
230]
. These contradictions
may be reconciled by a model according to which rewiring
does not occur at a constant rate, but rather in a rapid-then-
slow fashion
[55]
. Similar rapid-then-slow dynamics
characterize protein sequence divergence following dupli-
cation events, likely reflecting relaxed-then-tight selective
constraints on the function of the duplicated proteins.
Signatures of neo-functionalization, sub-functionalization
and asymmetric edge-specific divergence have been
observed
e
223,227
e
phenotype
relationships. In this chapter we have described the state-
of-the-art experimental and computational strategies
currently used to detect and predict binary and co-complex
protein interactions at proteome scale, and outlined the
major achievements of the field so far. We covered the new
concepts that will need to emerge, and the new technologies
that will need to be developed, so that complete reference
protein
cellular
systems
and
genotype
e
in
protein
protein
interactome
networks
protein interactome maps can materialize for
several organisms in the near future. With such maps in
hand, the principles governing interactome dynamics will
be deciphered and causal paths between genotype and
phenotype will be drawn.
e
e
[219,231
protein
interactome networks following duplication events thus
appears associated with the Darwinian selection of the
functions of the corresponding proteins.
233]
. Edgetic rewiring of protein
e
e
