Injectable RNA interference (RNAi) in adult Drosophila replicates a mutant phenotype.

Dzitoyeva, Svetlana, Nikola Dimitrijevic, and Hari Manev.  The Psychiatric Institute, Department of Psychiatry, University of Illinois at Chicago, 1601 West Taylor Street, MC912, Chicago, IL 60612, USA. E-mail: sdzitoyeva@psych.uic.edu

Introduction

      The discovery of RNA interference (RNAi) (for reviews, see Hannon, 2002;  Tijsterman et al., 2002) dramatically challenged and changed our approach to the study of functions of newly discovered genes.  In insects, the simple introduction of double-stranded RNA (dsRNA) corresponding to a particular gene either via direct injection into the embryo (Kennerdell and Carthew, 1998) and pupae (Bettencourt et al., 2002), or even into adult flies (Dzitoyeva et al., 2001) leads to the destruction of the corresponding endogenous mRNA, resulting in functional gene silencing. The physiological role of RNAi, particularly in the central nervous system (CNS), remains speculative (Smalheiser et al., 2001), but the applicability of experimentally induced RNAi for functional genomics, i.e. loss-of-fuction studies, is becoming more apparent (Gan et al., 2002).

      Several methods have been developed that allow for the creation of transgenic flies bearing “hairpin” (i.e., inverted repeat) constructs (Kennerdell and Carthew, 2000), which when expressed trigger RNAi.  Further improvements of this method include inverted repeats that form a double-stranded hairpin structure under the control of an upstream activating sequence (UAS) of the yeast transcriptional activator GAL4 (Piccin et al., 2001). Thus, an inducible and tissue-specific transgenic RNAi is also possible in Drosophila (Giordano et al., 2002). Nevertheless, a method of RNAi via injection of dsRNA may provide a simple and rapid alternative to these transgenic approaches.

      In our previous work, we demonstrated the ability of intra-abdominal dsRNA injections to induce RNAi in adult Drosophila. Hence, we observed silencing of mRNAs for several genes using dsRNAs corresponding to the entire cognate mRNA of the 3kb bacterial beta-galactosidase or using 1480bp of the CG2831 gene (AF145605) (Dzitoyeva et al., 2001). Here we describe a simple and straightforward procedure for injections of small 21-22 dsRNA molecules to trigger RNAi in adult Drosophila and thereby produce a phenotype that previously has been observed in mutant flies [i.e., fs(1)Yb]. An important event in Drosophila oogenesis is the division of the germline stem cell; mutations in fs(1)Yb (Yb) cause several defects in the adult ovary related to germline stem cell division (Johnson et al., 1995;  King and Lin, 1999).

Materials and Methods

      Synthesis of long dsRNA (>100 bp):  To synthesize RNA in vitro, plasmid DNA and RT-PCR templates could be used with SP6/ T7/ T3 RNA polymerases. Standard protocols for plasmid DNA-based RNA syntheses are used; e.g., those described in the Molecular Biology Protocols guide (http://highveld.com/ protocols.html). The reverse transcriptase-PCR based approach is relatively easy and allows for preparation of dsRNA probes without cloning of a specific fragment. One of the RNA polymerase promoter sequences T7/T3/SP6 is added to the 5’end of designed primers (better if different on each end), which facilitates the verification of the fragment orientation.

      For example, to demonstrate that endogenous mRNA can be destroyed in adult flies by injection of dsRNA, we targeted Indy (CG2831), a gene associated with longevity (Rogina et al., 2001). We designed the 38-mer primers with an attached T7 and T3 RNA polymerase promoter sequences to amplify the 1480 bp fragment of the CG2831 gene with a direct primer: 5’-taatacgact-cactatagggaattgcgtgtgcagaatg-3’ and reverse primer: 5’-taaccctcactaaagggatgcgtttcttgaggcttctc-3’. Standard RT-PCR reactions were used with the total extracted RNA from male and female flies. 0.1 µg of a PCR product was used to separately synthesize sense and antisense RNA strands. DNA template was removed by treating samples with RNAse-free DNAse (Ambion). Equal amounts of the single-stranded RNAs were mixed together, denatured by heating at 85ºC for 3-5 min and transferred to ice.  Thereafter, samples were extracted by phenol/chloroform (pH 4-5) and precipitated with ethanol. The quality of dsRNA was assessed on 1-1.5% of agarose gels; a good preparation shows a clear single band. If proportions of single-stranded RNAs are not equal, a smear will appear above and below the band of migrating dsRNA. Thereafter, we suggest the use of Enzyme Remover centrifuge columns before dsRNA is used for injections (check columns before use; they could bind your sample as well!).

      Synthesis of 21-22 bp dsRNA:  For experiments aimed at replicating a known mutant phenotype (King and Lin, 1999;   http://dev.biologists.org/cgi/reprint/126/9/1833.pdf) with an injection of small dsRNA molecules, we targeted a region (position 1142-1165) of the fs(1)Yb gene. As a control, we used dsRNA against the human 5-lipoxygenase (5-LOX) gene that has no known homologue in Drosophila. 38-39-mer DNA oligonucleotides with an attached T7 RNA polymerase promoter sequence representing sense and antisense strands were synthesized (Integrated DNA Tech. Inc.); 5’-3’ cDNA sequences were as follows:

      Fs(1)Yb:    5’-taatacgactcactatattttctgcagtgggaataactt-3’

      5-LOX:      5'-taatacgactcactataggtcattggacgagctgcctcc-3'

      A single match was found for the fs(1)Yb sequence (the genome databases pattern search analysis), ensuring the specificity of our tool for RNAi. Equal amounts of oligonucleotides were annealed at room temperature to form a double-stranded template. Standard in vitro transcription reaction was carried out. Sense and antisense samples were mixed together, denatured at 65ºC, and renatured at room temperature. The rest of the procedure was as described for the synthesis of longer dsRNAs (see above).

      Injections:  Wild type Drosophila melanogaster (CS/S) were cultured at 25°C, 50-60% humidity, 12 h/12 h light/dark cycle, on yeast, dark corn syrup, and agar food. Adult (i.e., 3-5 days old) male and female flies were used for experiments. Injections of 0.2µl/fly of dsRNA solutions were administered into the fly’s abdomen using a vacuum-operated fly-holder and a microinjector (for details, see Dzitoyeva et al., 2001).

Results and Discussion

Figure 1.  CG2831 dsRNA injection (80 ng/µl; 02 µl/fly) into adult flies destroys the endogenous CG2831 mRNA. The RT-PCR was used with total RNA extracted from female bodies after dsRNA injections: lane 1 = 24 h after CG2831 dsRNA, lane 2 = 48 h after CG2831 dsRNA, lane 3 = control. Lanes 4, 3, and 6 are the corresponding samples assayed for RP49 mRNA. Note the disappearance of CG2831 mRNA but not RP49 due to RNAi.

      We used the RT-PCR assay (Dzitoyeva et al., 2001) to assess the expression of CG2831 mRNA 24 and 48 hours after a control injection or after an injection of CG2831 dsRNA (Figure 1). The primers used in this assay were as follows: direct =5’gatatcggaattgcgtgtgcagaatg-3’, reverse = 5’-gatatctcaatggagctggt gaactc-3’. As an internal control in these experiments we assayed ribosomal gene RP49 mRNA, which was amplified using the following primers: direct = 5’-atgaccatccgcccagcataca-3’, reverse = 5’-tgtgtattccgaccaggttac-3’. No CG2831 PCR product was detected after injections of CG2831dsRNA, whereas these injections did not affect the RP49 PCR product (Figure 1).

      To demonstrate the ability of dsRNA injections into adult flies to replicate the morphological changes in ovaries observed in fs(1)Yb gene mutation (Johnson et al. 1995;  King and Lin, 1999), we injected dsRNA molecules designed against the position 1143-1165 of this gene. The fs(1)Yb mutant ovarioles typically contain two egg chambers (occasionally only one) and they have a small number of germline cells in the germarium (http://dev.biologists.org/cgi/reprint/126/9/1833.pdf). In about 70% of ovaries of flies injected with fs(1)Yb dsRNA we observed a Yb-like phenotype: i.e., very small and disorganized ovarioles containing one to three egg chambers, and relatively small embryos (Figure 2). As a control, we used dsRNA against the human 5-LOX; no visible changes were noticed after control dsRNA injections (Figure 2).

Figure 2.  Ovaries were isolated and fixed three days after injections of 50ng/µl of fs(1)Yb dsRNA or 5-LOX dsRNA (control):  A) control;  (B-D) RNAi-induced phenotype showing a small and disorganized ovary, and ovarioles containing one to three egg chambers.

      In the adult ovary, fs(1)Yb is expressed particularly in the terminal filament cells. The Yb gene is essential for the maintenance of germline stem cells (GSC) during oogenesis and plays an essential role in GSC self-renewal; loss of function leads to failure in GSC maintenance with severe phenotypic alterations. Our finding that the adult injectable RNAi is capable of replicating the Yb phenotype suggests that this gene is active and functionally important in the adult flies. Kennerdell et al. (2002) reported that gene silencing by dsRNA occurs in the Drosophila female germ line, but only against the transcripts that are translated. Thus, mRNA transcripts that are translationally quiescent (e.g., at the arrested oocyte stage) are insensitive to RNAi.

      In our initial work with injectable RNAi, as dsRNA we used the entire 3kb bacterial beta galactosidase construct, which is a part of the P-element transformation vector. The bacterial and the Drosophila beta galactosidase (also expressed endogenously; Schnetzer et al., 1996) nucleotide sequences share three exactly identical stretches, about 200 bp each. Apparently, this homology was sufficient to lead to the destruction of Drosophila’s own beta galactosidase mRNA by the bacterial dsRNA (Dzitoyeva et al., 2001). Thus, the homology between genes and their isoforms should be considered to avoid the possibility of any nonspecific inhibition by injected dsRNA. From our previous work, we realized that the concentration of injected dsRNA is critical for the induction of RNAi, and both in this earlier work and in this study we determined the working concentrations of our dsRNA probes empirically, making our observations consistent with those of others (Elbashir et al., 2001;  Zhou et al., 2002).

      Whereas the GAL4-driven hairpin-induced RNAi appears to be cell autonomous (Van Roessel et al., 2002), injecting adult Drosophila intra-abdominally with either long or short dsRNAs resulted in cell-nonautonomous silencing of the complementary endogenous mRNA and produced an altered phenotype. Thus, these two methods of RNAi induction in Drosophila could be used as complementary experimental approaches.

      In conclusion, we demonstrated that injecting adult Drosophila with either long or short dsRNA induces RNAi and leads to consecutive functional gene silencing that can be utilized in functional non-developmental characterizations of Drosophila genes.

      Acknowledgments:  This work was supported in part by the grant RO3DA14811 by the National Institute on Drug Abuse (H.M.).

            References:  Bettencourt, R, O. Terenius, and I. Faye 2002, Insect Mol. Biol. 11: 267-271;  Dzitoyeva, S., N. Dimitrijevic, and H. Manev 2001, Mol. Psychiatry 6: 665-670;  Elbashir, S.M.,  J. Harborth, W. Lendeckel, A. Yalcin, K. Weber, T. Tuschl 2001, Nature 411: 494-498;  Gan, L., K.E. Anton, B.A. Masterson, V.A.M. Vincent, S. Ye, and M. Gonzalez-Zulueta 2002, J. Neurosci. Methods 121: 151-157;  Giordano, E., R. Rendina, I. Peluso, and M. Furia 2002, Genetics 160: 637-648;  Hannon, G.J., 2002, Nature 418: 244-251;  Johnson, E., S. Wayne, and R. Nagoshi 1995, Genetics 140: 207-217;  Kennerdell, J.R., and R.W. Carthew 1998, Cell 95: 1017-1026;  Kennerdell, J.R., and R.W. Carthew 2000, Nat. Biotechnol. 17: 896-898;  Kennerdell, J.R., S. Yamaguchi, and R.W. Carthew 2002, Genes Dev. 16: 1884-1889;  King, F.J. and H. Lin 1999, Development 126: 1833-1844;  Piccin, A., A. Salameh, C. Benna, F. Sandrelli, G. Mazzotta, M. Zordan, E. Rosato, C.P. Rogina, B., R.A. Reenan, S.P. Nilsen, and S.L. Helfand 2000, Science 290: 2137-2140;  Kyriacou, and R. Costa 2001, Nucleic Acids Res. 29: E55-5;  Schnetzer, J.W., and M.S. Tyler 1996, Biol. Bull. 190: 173-187;  Smalheiser, N.R., H. Manev, and E. Costa 2001, Trends Neurosci. 24: 216-218;  Tijsterman, M., R.F. Ketting, and R.H. Plasterk 2002, Annu. Rev. Genet. 36: 489-519;  Van Roessel, P., N.M. Hayward, C.S. Barros, and A.H. Brand 2002, Genesis 34: 170-173;  Zhou, Y., Y.-P. Ching, K.H. Kok, H. Kung, and D.-Y. Jin 2002, Nucleic Acids Res. 30: 1664-1669.