Biology Reference
In-Depth Information
further analysis confirmed that this particular pathway (a
two-step biosynthetic route from aspartate to quinolinate)
was inferred erroneously from Escherichia coli [58] . This
study provides evidence for at least modest success of
mechanistic models to predict genetic interactions in vivo,
and perhaps more importantly, suggests a framework for
using targeted experimental measurements to refine exist-
ing metabolic network models. Such an approach could
have important applications as we begin to study genetic
interactions in higher eukaryotes, where experimental
technology for measuring genetic interactions is less scal-
able. In fact, similar metabolic network modeling
approaches have recently been applied successfully to
suggest new drug targets for specific cancers based on
predictions of synthetic lethal interactions [59,60] .
features, including cell number, nuclear area, and the
fluorescence intensity of stained nuclei [67] . Interestingly,
these phenotypic measurements were highly complemen-
tary and uncovered rich and non-redundant sets of genetic
interactions emphasizing the importance of diverse, quan-
titative phenotypic assays for mapping high-resolution
genetic networks comprehensively [67] (see Expanding
Genetic Networks: Mutant Alleles, Conditions and
Phenotypes). Consistent with yeast studies, genes that
shared similar function clustered together based on their
genetic interaction profiles, independent of the phenotypes
from which the profiles were generated. As a result, this
compendium of negative and positive interactions can be
used as a tool to predict Drosophila gene function. Indeed,
comparison of genetic interaction profiles led to the
discovery of a novel activator of RAS-MAPK pathway
signaling, whose function is conserved from fruit flies to
humans [67] . Importantly, although this study assayed
genetic interactions on an intermediate scale, its potential
for high-throughput application is obvious, suggesting that
genome-scale metazoan genetic networks, analogous to
those generated for yeast, are within grasp.
MAPPING GENETIC INTERACTIONS
IN OTHER ORGANISMS
Genetic Interactions in Unicellular
Organisms
Large-scale genetic interaction mapping techniques have
been developed for the fission yeast Schizosaccharomyces
pombe [61,62] and the Gram-negative bacterium E. coli
[63,64] . These techniques are directly analogous to those
employed in S. cerevisiae, in that they use genome-wide
deletion collections and mating procedures to generate
comprehensive sets of double mutants. Large-scale
mapping of S. pombe genetic interactions identified simi-
larities as well as significant differences in the wiring of its
genetic network compared to the S. cerevisiae network [61,
62] . Similar to metazoans, the S. pombe genome encodes
functional RNA interference (RNAi) machinery not found
in typical S. cerevisiae laboratory strains, which likely
account for at least some of the differences between the two
fungal genetic networks [62] .
Genetic Interactions in Mammalian Model
Systems
Currently, large-scale mammalian genetic interaction studies
are based on genetic manipulation of cell cultures via RNAi-
mediated gene inactivation. Approaches involving large
RNAi libraries and comparative analyses of different cell
lines have the potential to reveal genetic interactions specific
to cell line mutations. Indeed, several efforts focused on
specific cancer cell subtypes have discovered new synthetic
lethal interactions with cancer-associated genes, each of
which represents a new potential drug target. Two studies
have reported new synthetic lethal interactions of KRAS, the
serine/threonine kinase STK33 [68] and mitotic kinase PLK1
[69] , which have obvious therapeutic relevance given the
prevalence of KRAS mutations in various cancers. More
recently, several groups have identified common and cancer
cell-type-specific 'essential' genes in a panel of human cancer
cell lines [70,71] . One of these studies assessed the essenti-
ality of more than 11 000 genes in 102 human cancer cell
lines [70] . This large-scale effort uncovered 54 genes that are
specifically essential for the proliferation and viability of
ovarian cells, including PAX8 , which was identified as an
ovarian lineage-specific dependency [70] . These cell-type
specific interactions are presumably due to specific synthetic
lethal interactions between the RNAi target and the genetic
background of the tumor cell. Although the identification of
relevant endogenous second site mutation(s) remains to be
determined, a comprehensive understanding of the molecular
vulnerabilities of different cancer cell types will undoubtedly
provide a powerful roadmap to guide therapeutic approaches.
Genetic Interactions in Metazoan Model
Systems
Large-scale genetic interaction studies have also been
described for the nematode worm C. elegans [65,66] and
the fruit fly D. melanogaster [67] . These studies are made
possible by organism-specific genome-wide RNAi
libraries, which can be used to target individual genes and
are highly compatible with high-throughput experimental
approaches. A recent study combined a rigorous combi-
natorial RNAi experimental strategy with robust statistical
modeling to systematically map pairwise genetic interac-
tions between 93 genes encoding Drosophila signaling
factors [67] . Rather than focusing on a single phenotype,
identification of negative and positive genetic interactions
was based on measurement of three distinct phenotypic
Search WWH ::




Custom Search