Elsevier

Methods

Volume 68, Issue 1, 15 June 2014, Pages 82-88
Methods

RNAi screening in Drosophila cells and in vivo

https://doi.org/10.1016/j.ymeth.2014.02.018Get rights and content

Highlights

  • Reagents and assays for RNAi screens with Drosophila cells.

  • Fly stocks and assays for in vivo Drosophila RNAi screens.

  • Workflows for cell-based and in vivo RNAi screens.

  • Screen data analysis, enrichment and integration tools.

  • RNAi knockdown verification and ‘rescue’ approaches.

Abstract

Here, I discuss how RNAi screening can be used effectively to uncover gene function. Specifically, I discuss the types of high-throughput assays that can be done in Drosophila cells and in vivo, RNAi reagent design and available reagent collections, automated screen pipelines, analysis of screen results, and approaches to RNAi results verification.

Introduction

As a research tool, RNA interference (RNAi) harnesses an endogenous activity to reduce or “knock down” RNA levels in a sequence-specific manner. The availability of RNAi as a method has opened the door to functional genetic screening in new model systems and contexts (reviews include [1], [2], [3], [4], [5]). For Drosophila, RNAi made it possible to perform full-genome screens in cultured cells and in specific tissues in vivo using systematically designed and generated libraries of RNAi reagents, an important supplement to classical genetic screening [6]. For Drosophila cells, the active RNAi reagent is double-stranded RNA (dsRNA), typically 50–500 bp in length, which can be synthesized in vitro in 96-well format or larger volumes [7], [8]. For in vivo studies, transgenic fly stock collections were initially based on long dsRNA hairpin-encoding constructs introduced into flies via P-element transposon-based transgenesis [9]. The field subsequently moved to using short hairpin (shRNA)-encoding constructs in optimized vectors and introduced into flies using site-directed approaches, resulting in improved expression and knockdown, including in the germline [10].

RNAi is frequently used in Drosophila for low-throughput studies, such as when a large amount of relatively uniform material is needed (e.g. cell-based RNAi knockdown) or to study the effects of disruption of one or a few genes in a specific stage or tissue (e.g. using the Gal4-UAS system in vivo). RNAi is also contributing to development of Drosophila as a tool for personalized medicine as reviewed in [11]. One of the most powerful aspects of RNAi, however, is that it can be applied at genome scale. The focus of this chapter will be on large-scale studies. Nevertheless, many of the same approaches also apply to small-scale RNAi studies. In addition, many of the approaches presented here for RNAi screens are also relevant to other types of screens in cells or in vivo, e.g. over-expression [6] or small molecule screens. RNAi and small molecule screens can be performed in parallel as reviewed in [12] and exemplified by [13]. In the past, classical genetic and small molecule screening approaches helped inform best practices for RNAi screening. Similarly, we can expect that what we have learned from RNAi will influence approaches to new technologies for screening, e.g. genome engineering-based screening technologies [14].

Although the specific reagents and assays used for cell-based vs. in vivo RNAi screens are different, the two approaches share similarities in terms of information flow (Fig. 1). Before embarking on a screen on a particular topic, several decisions have to be made, including what assay to perform, what controls to use, and whether or not a full-genome library is appropriate – and if not, what subset of genes will be included in the screen. Researchers typically rely on pre-existing library collections for screens and thus, decisions regarding reagent design have largely been made for them. Nevertheless, reagent design, “coverage” (i.e. the number of unique reagents per gene), the frequency of updates to the library and other aspects of library design should be taken into consideration, as these will affect the number of plates or fly crosses in the screen, the quality and scope of results, and how data are analyzed later. Moreover, as gene annotations change over time, reagent annotations should be revisited prior to making final decisions about what set of reagents is used, as well as post-screening.

During the screen itself, a large number of samples and volume of data must be managed, with accurate preservation of the relationship between specific reagents, their positions on plates or in fly vials, and corresponding phenotype data. The leading edge in information tracking during an RNAi screen (or other large-scale screen, such as a small molecule screen) is tracking of screening plates or fly vials using barcode labels. For automated screens, barcode readers can be integrated into the automation pipeline. For fly vials handled directly by researchers, hand-held readers can be used to scan barcodes, followed by entry of phenotype data or other information into a spreadsheet or database. Following a large-scale image-based screen, the images must be analyzed, ideally using an automated approach. Images and other readouts are typically reduced to numerical scores, followed by statistical analysis to identify ‘hits’ in the primary screen. Next, the initial hits list is processed, additional experiments are performed, and together these results are used to define a high-confidence set of genes for further follow-up. Keeping careful digital records throughout a cell-based or in vivo screen is an essential basic task. For cell-based screens, researchers typically have access to data management tools through the library provider. For in vivo screens, researchers more typically set up their own databases or spreadsheets to track results. If phenotypes are expressed in text (e.g. “lethal” or “up-turned wing”) rather than auto-processed images or numbers, then standardized wording (i.e. a controlled vocabulary) should be established and used consistently in order to facilitate later analysis, interpretation and comparison.

Below I present more details on cell-based and in vivo RNAi screening, as well as information about data analysis and integration, reducing false positive discovery and experimental validation. In addition, I would like to point out that a large number of protocols, software tools, databases, helpful tips, stock collections, etc. relevant to Drosophila RNAi are freely available online. A few good starting points include the Drosophila RNAi Screening Center (DRSC) website (www.flyrnai.org) [15], the GenomeRNAi database (www.genomernai.org) [16], the Vienna Drosophila RNAi Center (VDRC) (http://stockcenter.vdrc.at/control/main), the National Institute of Genetics (NIG)-Fly (http://www.shigen.nig.ac.jp/fly/nigfly/), and the Sheffield RNAi Screening Facility (http://www.rnai.group.shef.ac.uk/). I also maintain an informational blog on Drosophila RNAi and related topics (http://flyrnai.blogspot.com/). Additional relevant online tools, resources and references are presented below and in Table 1, Table 2.

Section snippets

Overview of cell-based screens

More than a hundred Drosophila cell-based RNAi screens have been performed since the approach was first put to use at genome scale more than ten years ago [1]. Nevertheless, we have only scratched the surface in terms of the plethora of single- and multi-parameter assays that could be performed in this context (see Section 2.2. and 2.3). Cell-based screens in general are usually performed in one of two formats. For pooled screens, RNAi reagents are introduced at random, cells positive in the

Overview of Drosophila in vivo screening

Drosophila in vivo RNAi screens use the Gal4-UAS approach to drive expression of the RNAi reagent in a specific stage or tissue. Methods for Drosophila in vivo RNAi screening have been presented previously [3]. In addition, a broader discussion of genetic approaches is presented in another chapter [6]. The possibilities for in vivo RNAi assays are extremely broad, ranging from early developmental processes [45] to activities like flight that are restricted to adults [46]. The Gal4-UAS approach

General strategies for limiting false discovery

Large-scale studies are associated with reagent non-specific ‘noise’ that contributes to false discovery. Moreover, RNAi is associated with the possibility of reagent-specific off-target effects (OTEs), contributing to false positive discovery, as well as false negative results, e.g. due to a lack of robust knockdown. Others have previously reviewed or presented approaches to detecting and limiting the impact of false discovery, which include improvements to reagent design, improvements to

Verification of results from cell-based or in vivo RNAi screens

Verification of RNAi results is a significant bottleneck. Confirmation of knockdown efficiency via quantitative real-time PCR (qPCR) can provide supporting evidence for a relevant target gene-to-phenotype relationship. A method to assess protein depletion following RNAi in flies has also been reported [73]. Ultimately, however, verification of RNAi results rests on additional experimental analyses, such as rescue of the RNAi phenotype with an RNAi-resistant transgene and comparison of RNAi

Acknowledgements

I extend my thanks to Norbert Perrimon, Liz Perkins, and the DRSC/TRiP staff for many helpful discussions. The Drosophila RNAi Screening Center and the Transgenic RNAi Project at Harvard Medical School and are supported by NIH NIGMS R01 GM067761, NIGMS R01 GM084947 and NCRR/ORIP R24 RR032668 to N. Perrimon. I also receive support from the Dana Farber/Harvard Cancer Center, which is supported in part by NCI Cancer Center Support Grant # NIH 5 P30 CA06516 to E. Benz.

References (81)

  • N. Perrimon

    Drug Discovery Today

    (2007)
  • S. Armknecht

    Methods Enzymol.

    (2005)
  • S. Cherry

    Curr. Opin. Microbiol.

    (2008)
  • G. Goshima

    Methods Cell Biol.

    (2010)
  • D. Panda

    Cell Reports

    (2013)
  • J.H. Zhang et al.

    J. Biomol. Screen.

    (1999)
  • A.M. Wiles

    J. Biomol. Screen.

    (2008)
  • J. Marinho

    Develop. Biol.

    (2013)
  • R.A. Neumuller

    Cell Stem Cell

    (2011)
  • J.A. Pospisilik

    Cell

    (2010)
  • A. Jenett

    Cell Reports

    (2012)
  • S. Mohr et al.

    Annu. Rev. Biochem.

    (2010)
  • S.E. Mohr et al.

    Wiley Interdisciplinary Rev. RNA

    (2012)
  • N. Perrimon et al.

    Cold Spring Harbor Persp. Biol.

    (2010)
  • C. Bakal

    Briefings Funct. Genomics

    (2011)
  • C. Falschlehner

    Biotechnol. J.

    (2010)
  • K. Venken, H. Bellen, Chemical Mutagens, Transposons, and Transgenes to Interrogate Gene Function in Drosophila...
  • N. Ramadan

    Nat. Protoc.

    (2007)
  • S. Steinbrink et al.

    Methods Mol. Biol.

    (2008)
  • G. Dietzl

    Nature

    (2007)
  • J.Q. Ni

    Nat. Methods

    (2011)
  • Y. Kasai et al.

    Personalized Med.

    (2010)
  • J. Schulte

    PLoS ONE

    (2011)
  • K.J. Beumer, D. Carroll, Targeted genome engineering techniques in Drosophila, Methods (2014) METHODS-D-13-00261...
  • I.T. Flockhart

    Nucleic Acids Res.

    (2012)
  • E.E. Schmidt

    Nucleic Acids Res.

    (2013)
  • B. Baum et al.

    Methods Mol. Biol.

    (2008)
  • L. Cherbas, L. Gong, Cell lines. Methods 68 (1) (2014)...
  • L. Cherbas

    Genome Res.

    (2011)
  • A. Simcox

    PLoS Genet.

    (2008)
  • N. Perrimon et al.

    J. Visual. Exp. (JoVE)

    (2011)
  • J. Bai et al.

    Nat. Protoc.

    (2009)
  • N. Perrimon et al.

    Genetics

    (2007)
  • R. DasGupta

    Genome Biol.

    (2007)
  • M. Bettencourt-Dias et al.

    Methods Mol. Biol.

    (2009)
  • R. DasGupta et al.

    Methods Mol. Biol.

    (2008)
  • R.A. Neumuller

    Sci. Signal.

    (2013)
  • E.F. Joyce

    PLoS Genet.

    (2012)
  • R. Aikin

    PLoS ONE

    (2012)
  • T. Horn

    Nat. Methods

    (2011)
  • Cited by (29)

    • Genetics in Light of Transcriptional Adaptation

      2020, Trends in Genetics
      Citation Excerpt :

      However, unlike small molecules, little prior knowledge other than gene sequence is necessary for the design and use of antisense approaches. Thus, their versatility made them very popular with animal models, especially where efficient gene targeting was not available, including worms, flies, fish, frogs, and chickens [12–14]. Together, these functional studies increased our understanding of the role of different factors during development and organogenesis.

    • Daywake, an Anti-siesta Gene Linked to a Splicing-Based Thermostat from an Adjoining Clock Gene

      2019, Current Biology
      Citation Excerpt :

      Since increases in dmpi8 splicing efficiency are causally linked to lower midday siesta levels [7], a testable prediction from our findings is that downregulating 0.9 expression should increase daytime sleep levels. To this end, we used the Gal4-UAS system [29] in combination with RNAi [30] against 0.9 (RNAi-0.9) to more directly test for a role for 0.9 in wake-sleep regulation. In strong agreement with our hypothesis, driving RNAi-0.9 in dper-expressing cells markedly increased daytime sleep levels relative to both parental control crosses, whereas nighttime sleep levels exhibited little change with this treatment (Figures 3A–3E and S3).

    • Functional Genomics Screens in Drosophila Cells

      2018, Drosophila Cells in Culture: Second Edition
    View all citing articles on Scopus
    View full text