RNAi Methods and Screening: RNAi Based High-Throughput Genetic Interaction Screening - Caenorhabditis Elegans: Molecular Genetics and Development - page 91

Davierwala et al., 2005; Tong et al.,2004 ) and has begun in C. elegans ( Byrne et al.,
2007; Holway et al., 2005; Lehner et al., 2006; Tischler et al., 2006 ).
Multiple approaches will be needed to map all genetic interactions in a complex
multi-cellular organism, including ways that rely on dramatically expanded abilities
to perform, archive, and analyze ultra-large systematic screens. Here we explore
double genetic combinations using a method to test RNAi effects of single genes in
different genetic backgrounds.
In C. elegans, mutants affecting most cellular processes are available from a
public repository, the Caenorhabditis Genetics Center (CGC). In addition, there
are several ongoing projects to produce mutants targeting almost every gene in
the genome, including the C. elegans Knockout Consortium ( http://celeganskocon-
sortium.omrf.org/ ) , the C. elegans National Bioresource Project of Japan ( http://
shigen.lab.nig.ac.jp/c.elegans/index.jsp ), and the NemaGENETAG project ( http://
elegans.imbb.forth.gr/nemagenetag/ ) .
Early large-scale RNAi studies concentrated on individual gene analyses
( Fernandez et al., 2005; Fraser et al., 2000; Gonczy et al., 2000; Gunsalus
and Piano, 2005; Kamath et al., 2003; Maeda and Sugimoto, 2001; Piano
et al., 2000; Rual et al., 2004; Sonnichsen et al., 2005 ). A result from these
studies was that less than 20% of genes produced a clear phenotype under
laboratory conditions. Although this observation could be due to a number of
reasons, including incomplete functional depletion using RNAi or not scoring
for all phenotypes, the effect of losing one gene can be masked by genetic
redundancy or other compensatory mechanisms. Since genes interact with one
another to modulate cellular systems and generate specific phenotypes, it is
useful to develop methods that reveal genetic interaction networks in a multi-
cellular organism in large scale.
Amajor technical bottleneck of increasing the throughput of large-scale screening
has been the limited window of time in which the results need to be scored. In the
following, we describe the pipeline that we have implemented in our lab to overcome
this issue. We modified a liquid RNAi protocol in 96-well plates ( Ahringer, 2006 )to
use with temperature-sensitive (ts) mutants, and we use it to perform 40,000 RNAi
experiments per 9-day cycle. In addition, we developed an image-capturing platform
that quickly records and archives high-quality images of all the experiments. This
has given us the ability to separate the experimental testing from the analysis of the
data. We can score data long after it is collected and review it multiple times.
Moreover, this approach has opened the door to using computer vision to help
develop automation and quantitative data analysis for the phenotypic scoring
( White et al., 2010 ).
We implemented these methods in screens using conditional mutations (e.g.,
temperature-sensitive), enabling us to control the timing of when we assay for
genetic interactions (our unpublished data). Since our studies focus on embryonic
development, we maintain the larvae at permissive temperature and do not perturb
the gene function until the last larval stage (L4) to bypass genetic interactions that
could lead to adult phenotypes before embryos are produced.