Drosophila as a Genetic Model System for Molecular Neurobiology

The entire genome of Drosophila has been sequenced (Adams et al., 2000), and EST cDNAs are available from multiple libraries representing different stages of development (BDGP, http://www.fruitfly.org/). The availability of both genomic sequence and cDNA clones has increased the efficiency of forward and reverse genetic approaches (for review, see Adams and Sekelsky, 2002). Mutant screens for novel genes remain a primary tool of Drosophila neurobiologists (for examples of reviews, see Broadie, 1998; Garcia-Alonso, 1999; Wager-Smith and Kay, 2000; Featherstone and Broadie, 2000; Waddell and Quin, 2000; Sokolowski, 2001; Jin, 2002; for general review about the methods and rationale of Drosophila screens, see St. Johnston, 2002). Screens involve mutagenizing the genome with transpos-able elements (e.g., P-element), chemical mutagens or ionizing radiation, and screening for desired phenotypes. A new method of screening is through the use of a double element consisting of a transposable element (Hobo) bracketed by two genetic markers all in a "carrier" P element (Huet et al., 2002). Once in flies, it can be mobilized to make nested deletions, and is desirable because it does not suffer from the site specificity of P-element insertion (Spradling et al., 1999). Comprehensive collections of deletions spanning the genome are continually being developed for mapping genomic regions and to allow a focus on smaller regions of interest (Huet et al., 2002).

P-element-generated mutants allow one to rapidly identify the relevant gene, as many versions of P-elements have been designed to make it relatively simple to recover the genomic sequences flanking the site of insertion [for details, see the two chapters by Wolfner and Goldberg (1994) and by Hamiliton and Zinn (1994), a chapter by Huang et al., (2000), sections of the book by R.J. Greenspan, Adams, and Sekelsky, (2002) and the BDGP, http://www.fruitfly.org/]. However, new methods for identifying the site of chemically induced mutations, which are less time-consuming, are currently being refined (Bentley et al., 2000; Teeter et al., 2000; Hoskins et al., 2001; Berger et al., 2001). A recently developed method that is particularly sensitive, but relatively expensive, is called denaturing high-performance liquid chromatography (DHPLC) and is based on polymerase chain reaction (PCR) analyses of heterozygous flies leading to both homoduplexes and hetero-duplexes of the mutagenized regions (Bentley et al., 2000). The heteroduplexes have altered heat stability and are separated on an HPLC column at the temperature causing partial denaturation. This method is reliable but requires a large investment of equipment. Another method of mapping mutants is called single nucleotide polymorphism (SNP) mapping, and a description of the project developing this method may be found at BDGP (http://www.fruitfly.org/SNP/ index.html). This method is in the process of rapid development in Drosophila (Teeter et al., 2000; Hoskins et al., 2001; Berger et al., 2001; Martin et al., 2001) and relies on single base pair polymorphisms between specific laboratory strains of flies.

A. Rapid Reverse Genetic Approaches for Initial Assessment of Gene Function

Although forward genetic screens are a real power of Drosophila, the detailed knowledge of the genome is rapidly increasing the need for and use of reverse genetic methods to test the function of genes previously identified in vertebrates or through the analysis of human diseases (for review, see Link, 2001). Several approaches can be used to determine if a target gene is involved in a process of interest. First, one may look to see if a mutant strain already exists by searching available stocks at Drosophila stock centers (http://flybase.bio.indiana.edu/stocks/ ). Because cDNAs can be obtained (or generated) easily, in situ RNA hybridization (Tautz and Pfeifle, 1989; Lehman and Tautz, 1994) will quickly tell one where and when the gene is expressed in the nervous system. Second, a tool recently developed to rapidly assess the function of a gene is RNA interference (RNAi) (Fire et al., 1999; Kennerdell and Carthew, 1998; Misquitta and Patterson, 1999; Li et al., 2000; Wianny and Zernicka-Goetz, 2000; for review, see Sharp, 2001; Fjose et al., 2001; Schmid et al., 2002; Hutvagner and Zamore, 2002). RNAi effectively silences genes at a posttranscriptional stage. Application of double-stranded RNA (dsRNA) corresponding to a single gene leads to the enzymatic degradation of the relevant mRNA having the same sequence. This technique was first used in C. elegans and more recently in several other organisms (Fire et al., 1999; Kennerdell and Carthew, 1998; Li et al., 2000; Wianny and Zernicka-Goetz, 2000). Bathing (Eaton et al., 2002) or injection of double-stranded RNA into embryos (Misquitta and Patterson, 1999) or even adult abdomens (Dzitoyeva et al., 2001) can quickly give preliminary data indicating the likelihood of a gene playing a neurobiological role. RNAi can also be used via transgenic methods; however, it takes months to create the flies and is not always as reliable as classical mutants, but it does allow for tissue-specific RNAi dissection (Lam and Thummel, 2000; Kennerdell and Carthew, 2000; Fortier and Belote, 2000; Piccin et al., 2001; Allikian et al., 2002; Kalidas and Smith, 2002; see later).

A chapter outlining the technique of RNAi injection can be found in Misquitta and Patterson (2000). An example of a test of the rapid injection approach has been described by Schmidt et al. (2002). In this study, the role of five neural receptor tyrosine phosphatases (RPTP) in axon development was tested by injecting various combinations of double-stranded RNAs into embryos and subsequent injection of a lipophilic dye (Dil) into identified neuroblasts. This allowed visualization of identified neurons derived from known progenitors and the effects of RPTP RNAi on these cells. These experiments were particularly beneficial because mutations in three of the five RPTPs are embryonic lethal and the other two show no phenotype. This method allowed them to easily address the issue of functional redundancy. Using RNAi, quadruple mutants were generated, allowing the demonstration that all four RPTPs play a role in axon guidance. In this report, it was clear that identifying the correct dose of RNAi to yield high efficiency of gene silencing and at the same time avoiding nonspecific toxic effects is not trivial. Additionally, this method relies on protein turnover to approach a "protein null'' phenotype at the time when one observes the RNAi-treated cell. For these reasons, among others, the next step after this type of RNAi experiment should be to generate or obtain mutants of the relevant gene(s).

More recently, several laboratories have used RNAi by transgenic methods to assay function for a given gene because injected dsRNA does not interfere reliably with gene function later in development (Lam and Thummel, 2000; Kennerdell and Carthew, 2000; Fortier and Belote, 2000; Piccin et al., 2001; Allikian et al., 2002; Kalidas and Smith, 2002). This allows for efficient targeting of RNAi in adult tissues. This new method depends on the creation of transgenic flies carrying constructs that drive the transcription of RNAs with inverted repeats separated by a linker, such as an intron. These RNAs then fold back on themselves to form dsRNA molecules. Expression of the dsRNA can be driven by the Gal4UAS system described later. The Gal4UAS system drives expression in a tissue-specific manner, and recent permutations allow for finer temporal regulation of expression (Bieschke etal., 1998; Bello etal., 1998; Osterwalder etal., 2001; Roman etal., 2001; Stebbins and Yin, 2001; Stebbins etal., 2001; Landis etal., 2001). These newer methods rely on the control of transcription by agents fed to flies, such as RU486 or doxycycline, and therefore allows for the timing of RNAi expression to be controlled temporally and spatially in a dose-dependent manner (Allikian et al., 2002).

B. Generation of Mutations in a Gene of Interest

There are several standard methods of generating mutants, each with its own advantages and disadvantages. Many P-element collections exist, and information regarding these lines is available via a searchable database at BDGP [for a review of methods, see chapter by Mistra et al., (2000) and review by Adams and Sekelsky (2002)]. Standard methods can be used to mobilize a P-element into your gene or generate precise/imprecise P-element excisions [for details, see the two chapters by Wolfner and Goldberg (2000), by Hamilton and Zinn (2000), sections in R. J. Greenspan's book and Adams and Sekelsky (2002)]. However, occasionally the insertion of a P-element into a particular gene occurs with very low frequency because P-elements display a bias for certain sequences (Spradling et al., 1999). In the event that this occurs, larger deletions can be made from the sites of nearby insertions (Adams and Sekelsky, 2002), but the risk is that other genes are likely to be deleted. Precise excisions restore the original sequence and are critical because they demonstrate that the observed phenotype is due to the P-element insertion; imprecise excisions generate mutants ranging from nulls to hypomorphs (reduced function). A note of warning: one has to be careful with any P-element line, as a P-element can be mobilized more than once in the germ line, thereby disrupting the function of another gene that may not be easy to identify. Thus, the best policy is to outcross the mutant line with a wild-type reference line that should eliminate second site hits. Generally, P-elements insertions lead to hypomorphs or mutants with low gene activity because of the tendancy of P-elements to insert into 5' regions, leading to reduced expression (Spradling et al., 1999). Typically, imprecise excisions of P-elements result in null alleles.

Chemical mutagen-induced point mutations are more likely to generate informative hypomorphs and null alleles. Classical methods of mutagenesis in flies have relied on either chemical mutagens or ionizing radiation [for radiation methods, see Sankaranarayanan and Sobels (1976) and Grigliatti (1998); for chemical mutagens, see Lee (1976), Grigliatti (1998), and Adams and Sekelsky (2002)]. Chemical mutagens [ethyl methanesulfonate (EMS) or, more recently, N-ethyl-N-nitrosourea (ENU)] tend to give point mutations, whereas irradiation gives deletions and other chromosomal aberrations. ENU (Vogel and Natarajan, 1979; Batzer et al., 1998; Lee et al., 1990) has gained popularity because it produces fewer translocations and appears to be more potent than EMS (Vogel and Natarajan, 1979; Lee et al., 1990). Both EMS and ENU are very useful because these mutagens yield high mutation rates with a high probability of hitting any given gene. Therefore, the frequency of mutagenizing any particular gene depends essentially on the size of the gene, unlike P-elements, which insert into distinct "hot spots'' (Spradling et al., 1999). Chemical mutagens generally do lead to mutations at specific bases, but these bases are found in all genes. Another advantage of chemical mutagens is that one can potentially identify unknown critical domains in a protein by identifying point mutations by sequence analysis. Ionizing radiation produces a full spectrum of mutations, including deletions (Sankaranarayanan and Sobels, 1976; Grigliatti, 1998). It might be best used as a method of generating deletions in a gene in which it is difficult to obtain a P-element in the region.

Another method of generating a mutation in a gene of choice is by using random mutagenesis on flies heterozygous for a deficiency spanning the region covering the gene and then screening for a predicted phenotype. Flies have been treated by various means leading to chromosomal abberations, including those with deletions of portions of chromosomes (deficiencies, Gubb, 1998). Flies carrying deficiencies covering the genome (deficiency kit) can be found by searching Flybase and are available from the Bloomington stock (http://flystocks. bio.indiana.edu/). These methods using deletions rely on being able to predict the mutant phenotype (Saxton et al., 1991; Renden et al., 2001) or detecting the loss of or modification of the protein by Western blotting (Katz et al., 1988; Van Vactor et al., 1988; Dolph et al., 1993).

Was this article helpful?

0 0

Post a comment