Elsevier

Methods

Volume 30, Issue 4, August 2003, Pages 313-321
Methods

Genome-wide RNAi screening in Caenorhabditis elegans

https://doi.org/10.1016/S1046-2023(03)00050-1Get rights and content

Abstract

In Caenorhabditis elegans, introduction of double-stranded RNA (dsRNA) results in the specific inactivation of an endogenous gene with corresponding sequence; this technique is known as RNA interference (RNAi). It has previously been shown that RNAi can be performed by direct microinjection of dsRNA into adult hermaphrodite worms, by soaking worms in a solution of dsRNA, or by feeding worms Escherichia coli expressing target-gene dsRNA. We have developed a simple optimized protocol exploiting this third mode of dsRNA introduction, RNAi by feeding, which allows rapid and effective analysis of gene function in C. elegans. Furthermore, we have constructed a library of bacterial strains corresponding to roughly 86% of the estimated 19,000 predicted genes in C. elegans, and we have used it to perform genome-wide analyses of gene function. This library is publicly available, reusable resource allowing for rapid large-scale RNAi experiments. We have used this library to perform genome-wide analyses of gene function in C. elegans. Here, we describe the protocols used for bacterial library construction and for high-throughput screening in C. elegans using RNAi by feeding.

Introduction

Because of its anatomic and genetic simplicity, the nematode Caenorhabditis elegans has become an intensively studied simple animal model system. Furthermore, in 1998, C. elegans was the first multicellular eukaryote to have its genome completely sequenced, revealing 97 Mb of sequence encoding approximately 19,000 predicted genes [1]. The majority of these predicted genes remain uncharacterized; thus, although the genotype and phenotype of C. elegans have been studied in exquisite detail, the manner in which its genomic sequence is able to ultimately specify phenotype remains unclear. The study of genetics in C. elegans, from its inception as an experimental organism, has been largely based on classical forward genetics. Saturation forward-genetic screens have been extremely successful in comprehensively defining gene functions in organisms with compact genomes; however, this approach is less amenable to larger eukaryotic genomes since the sheer amount of time required to screen for and map every mutable gene is prohibitive. Completion of the C. elegans genome sequence, however, has greatly enhanced the prospect of employing reverse genetics as a complementary approach in the identification of gene function on a large scale.

Reverse genetics in C. elegans has been much more prevalent since 1998, when it was shown that the introduction of double-stranded RNA into a hermaphrodite worm results in potent and specific inactivation of an endogenous gene with corresponding sequence [2]. This technique, known as RNA interference (RNAi), enables rapid, targeted gene inactivation and has become an extremely important tool for studying gene function in vivo; moreover, because it is the simplest and quickest means of inactivating genes in C. elegans, RNAi has been rapidly embraced as a reverse-genetic tool and has dramatically accelerated the pace at which new gene functions are discovered. Initial studies of RNAi showed that injection of dsRNA into the head or tail of the animal was able to produce robust interference throughout the injected animal, including the germ line, suggesting that RNAi has the capacity to cross cellular boundaries. Indeed, subsequent studies demonstrated that RNAi could be performed in C. elegans simply by soaking worms in a solution of dsRNA [3] or by feeding them Escherichia coli expressing target-gene dsRNA [4] and that RNAi by feeding is an effective screening tool to determine the loss-of-function phenotype of a gene of interest [5], [6].

All three methods (injection, soaking, and feeding) have been used effectively in large-scale RNAi screening [7], [8], [9], [10]. Of these, RNAi by feeding has many advantages. First, because feeding is far less labor-intensive than microinjection, it is convenient for performing RNAi on a large number of worms or testing a large number of different genes. Second, feeding is considerably less expensive than either injection or soaking, which require the in vitro synthesis of dsRNA. However, RNAi by feeding suffers from one major drawback limiting its use as a high-throughput functional genomic tool: for each gene tested, it requires significant molecular biology work to clone a DNA fragment from the gene of interest into a special plasmid vector and then to transform it into a particular bacterial strain. To circumvent this limitation, we have constructed a bacterial library corresponding to roughly 86% of the estimated 19,000 predicted genes in C. elegans [7], [11]. This library has already been screened for genes involved in a wide range of biological processes, including embryonic development [12], aging [13], [14], fat regulation [15], and genome stability [16], and it should facilitate the ability of other investigators to perform their own genome-wide RNAi screens. Here we describe effective high-throughput methods for RNAi by feeding on a large scale. In addition, as the existing RNAi feeding library is not complete, we also describe methods for bacterial feeding library construction. Use of these methods and reagents should accelerate the identification of new gene functions in C. elegans.

Section snippets

Background information

The sheer ease and speed of RNAi as an experimental technique have made it feasible to rapidly determine the loss-of-function phenotypes of large numbers of genes. Shortly after the C. elegans genome sequence became available, primers were designed that were capable of amplifying PCR products from each C. elegans predicted gene for use in generating DNA microarrays (S. Jones, personal communication); these primers (GenePairs) were synthesized and are commercially distributed by Research

Discussion

RNAi is a remarkably convenient method for determining the loss-of-function phenotype of an individual gene of interest. By performing RNAi on a large scale using a genome-wide RNAi library, this reverse-genetic method can essentially be used as a forward-genetic screening tool. In performing such a large-scale RNAi screen, several important factors must be considered. First, genes have different sensitivities to RNAi. Genes that encode proteins with long half-lives can be difficult to target,

Acknowledgements

We are grateful to Andrew Fraser for critically reading the manuscript. R.S.K. was supported by a Howard Hughes Medical Institute Predoctoral Fellowship and J.A. by a Wellcome Trust Senior Research Fellowship (No. 054523).

References (32)

  • L. Timmons et al.

    Gene

    (2001)
  • I. Maeda et al.

    Curr. Biol.

    (2001)
  • F. Piano et al.

    Curr. Biol.

    (2000)
  • H. Inoue et al.

    Gene

    (1990)
  • T. Sijen et al.

    Cell

    (2001)
  • F. Simmer et al.

    Curr. Biol.

    (2002)
  • C. elegans Sequencing Consortium, Science 282 (1998)...
  • A. Fire et al.

    Nature

    (1998)
  • H. Tabara et al.

    Science

    (1998)
  • L. Timmons et al.

    Nature

    (1998)
  • R.S. Kamath et al.

    Genome Biol.

    (2001)
  • A.G. Fraser et al.

    Nature

    (2000)
  • P. Gonczy et al.

    Nature

    (2000)
  • R.S. Kamath et al.

    Nature

    (2003)
  • P. Zipperlen et al.

    Embo. J.

    (2001)
  • A. Dillin et al.

    Science

    (2002)
  • Cited by (1088)

    View all citing articles on Scopus
    View full text