Cell Biology Techniques in the Drosophila Nervous System

Among the forward genetic systems, methods for the analysis of neurons in situ are the most advanced in Drosophila. Immunocytochemical methods are developed for all stages of the life cycle and many antibodies are available (see Table II). In C. elegans, creating transgenic animals expressing marked proteins is much faster, but regulating the expression levels is fairly difficult. In flies, however, transgenic protein expression is a standard method and many variations of controlling expression can be used. Genomic (Spradling, 1986) or ectopic expression (Brand and Perrimon, 1993; Brand et al., 1994; van Roessel and Brand, 2002) is routine, and inducible systems have also been introduced (Bieschke et al., 1998; Bello et al., 1998; Osterwalder et al., 2001; Roman et al., 2001; Stebbins and Yin, 2001; Stebbins et al., 2001; Landis et al., 2001). Newly developed gene targeting methods should also allow "knockins" of tags and mutations (Rong and Golic, 2000, 2001; Bibikova et al., 2002), leading to mutated proteins expressed under the control of the endogenous genomic promoter. Finally, in flies it is possible to study the phenotypes of early lethal mutations by several clonal analysis techniques in neural tissues. Before these methods are discussed, the Drosophila life cycle is outlined briefly.

A. The Drosophila Life Cycle

During its holometabolous life cycle, Drosophila passes through several developmental stages, each of which has distinct advantages (and disadvantages) for a neurobiologist. The development of Drosophila is highly temperature dependent. The times given here represent development at 25° C, at which eggs are laid under optimal conditions. The eggs are laid soon after fertilization and subsequently undergo 17 defined embryonic stages, together lasting 20-22 h. During this time embryos develop and ultimately hatch into first instar (L1) larvae. The larvae undergo two molts, leading to a succession of three larval instar states (L1, L2, and L3) lasting ~96h. Mature third instar larvae enter a "wandering stage,'' leaving the food in search of a suitable site for pupariation. Adults eclose from the pupal case after 4-4.5 days. Adult flies typically live 30-60 days. For an account of the development of the structure and electrical properties of neurons during development, see Budnik and Gramates (1999).

B. Protein Expression Studies with Immunocytochemistry

Many antibodies are available to probe identified classes of Drosophila neurons and glia in the embryo, larva, and adult. Many of these antibodies may be used to identify subcellular compartments within neurons or glia. A list of available antibodies is provided in Table II. Many of these antibodies are available from the Developmental Studies Hybridoma Bank at the University of Iowa (DSHB: http://www.uiowa.edu/~dshbwww/). Because fixation and staining protocols must be worked out empirically for any given antibody, a simple set of immunocyto-chemistry instructions is not feasible. Variations in fix type/duration, dilutions, and other specific protocols may be found in the original papers, as indicated in Table II.

Method A: Immunolabeling the Embryonic Nervous System

For optimal egg collection, maintain young (<7 days), healthy male and female flies (20-40) on fresh laying pots for 2-3 days. Laying pots consist of 100-ml plastic beakers (Tri-Pour) perforated to provide adequate air circulation, covering a 60-mm agar plate. Agar plates contain apple or grape juice hardened with agar. Several recipes exist, but one example is as follows: 700 ml of water, 25-30 g agar, 300 ml of juice concentrate (grape or apple), 0.5 g methyl paraben (p-hydroxymethylbenzoate), and 30 g sugar. Autoclave the water and agar and, separately, boil the p-hydroxymethylbenzoate with the sugar and juice. Mix the agar and juice solutions and quickly pour into 60-mm plates, removing bubbles. Prior to egg collection, flies are fed daily with yeast paste (7 g baker's yeast in 9 ml of water stored at 4° C) on fresh plates twice a day. By day 3, a good pot will produce 100-200 eggs per hour. Better egg laying is observed on a pot maintained on its side with the agar in the plate scratched. Many embryos will be laid in or next to the scratches.

For embryo collection, gently transfer the embryos from the juice plate and into a basket using a paintbrush. A basket can be made with a 15-or 50-ml centrifuge tube. Cut off the bottom, leaving enough surface area to trap a screen (70 ym mesh) between the lid and the top of the tube. Rinse the embryos with dH2O and put into a petri dish of fresh 50 to 100% bleach to remove the outer chorion. Dechorionation can take from 30 s to 2 min and should be monitored under the microscope. Removal of the chorion exposes the shiny, transparent vitelline membrane. The next step is to permeabilize the vitelline membrane, which is achieved simultaneously with fixation. Fixation protocols may vary for different antibodies. For a standard fixation proceedure, place 2 ml of heptane and 2 ml of 4% paraformaldehyde fix in a glass vial (paraformaldehyde is best made fresh, but for routine experiments we store 5-ml aliquots in the freezer). Drop the embryo basket into the vial, shake off the embryos, and remove the screen. Fix for 20 min at room temperature with agitation. Remove the lower phase containing the paraformaldehyde and add 2 ml 100% methanol. Shake vigorously for 30 s to rupture the vitelline membrane. Allow embryos to settle (only the devitellinized embryos sink to the bottom of the vial). Aspirate the upper heptane phase and remove most of the methanol, and wash embryos three times with methanol. At this point, embryos can be stored in methanol for many weeks to months at —20° C.

Antibody staining protocols vary, but a standard approach is given here. Transfer embryos to a microcentrifuge tube and rehydrate in 50% methanol. Rinse three times in PBS-TX (0.02M phosphate buffer and 0.1 M NaCl, pH 7, 0.1% Triton X-100; note that detergent use can also vary with different antibodies) and three more times in PBS-T-BSA (PBS, 0.1% Tween 20, 4% BSA). Block by incubating for 1 h in PBS-T-BSA. After 1 h, remove PBS-T-BSA and add antibodies diluted in PBS-T-BSA. Incubate at least 2h at room temperature or overnight at 4° C. Allow embryos to settle, and remove and save antibody by adding 0.01% azide. For routine studies, these antibodies can generally be used five times if quantitation is not the main issue being examined. Rinse embryos four times in PBS-T-BSA over a period of 1 h. Add an appropriate dilution of secondary antibody conjugated to a flourescent chromophore, an enzyme (HRP), or biotin. For fluorescent secondary antibodies, dilute (typically 1:100-500) in PBS-T-BSA and incubate 1-2 h at room temperature in tubes covered with aluminum foil. Wash twice with PBS-T-BSA for 30 min and four times with PBS over a period of 30 min. Mount in a protective medium (e.g., Vectorshield) and observe by light or confocal microscopy.

For HRP-conjugated antibodies or biotin-labeled secondary antibodies, dilute (typically 1:100-500) in PBS-T-BSA and incubate for 1-2 h or overnight at 4° C. Wash the embryos twice with PBS-T-BSA for 30 min and four times with PBS over a period of 30 min. If amplification via the avidin/biotin-HRP system is used (Vectastain kit, Vector Laboratories), prepare a 1:200 dilution of avidin (reagent A) and biotinylated-HRP (reagent B) in PBS-T-BSA. Incubate the embryos in this mixture at room temperature for 1 h. Wash the embryos twice with PBS-T-BSA for 30 min and four times with PBS-TX over a period of 30 min. Begin developing in 600 pl PBS-TX by adding 20 pl of 2% NiCl2, and then bring the sample to 0.06% H2O2 using a freshly made 30% stock of H2O2. Put embryos into a watchglass and add 30 ^l of 10mg/ml diaminobenzidene (DAB). One at a time, add to each sample 20 ^l of 1:500 H2O2, swirl, and watch reaction in the microscope. Wait until the NMJ becomes visible and then stop the reaction by replacing buffer with fresh PBS-TX. Rinse three times with PBS-TX. Prepare samples in the manner allowing one to mount the samples in your favorite permanent mounting medium (e.g., Permamount).

Method B: Embryonic and Larval Dissection for Immunolabeling

Embryos can be dechorionated with bleach and devitellinated manually, and the developmental stage determined either temporally or using morphological criteria. Homozygous mutant embryos can be selected from siblings based on the absence of balancer chromosome markers (e.g., yellow, GFP). Balancer chromosomes are dominantly marked, multiply inverted chromosomes, which allow one to maintain heterozygous stocks of lethal mutants (Greenspan, 1997). This tool is another strength of Drosophila that makes maintaining stocks and generating crosses simple. For embryonic lethal mutations, one-quarter of the eggs are homozygotes, one-quarter will have two balancer chromosomes, and half will be heterozygotes. Embryos lacking the balancer will be unmarked. The most readily scored marker is the green fluorescent protein (GFP) (Tsien, 1998), and balancer chromosomes marked with GFP can be obtained from the Bloomington stock center.

For dissection, it works best if staged embryos are glued (Histoacryl Blue, Braun Germany) to Sylgard-coated coverslips under physiological saline (Sylgard resin from Dow Corning, Corning, NY). The saline contains (in mM): 135 NaCl, 5 KCl, 4 MgCl2, 1.8 CaCl2, 72 sucrose, 5 TES, pH 7.2; however, alternative salines are also used (for a complete list of salines and comparison with hemolymph measurements, see Broadie, 2000a). The addition of sucrose approximates the osmotic strength reported in a natural hemolymph (Stewart et al., 1994). Following gluing of the head and tail, a slit is made along either the dorsal or the ventral midline, using a glass capillary pulled to a sharp point or metal needle. The body walls are then glued flat to the coverslip, and the internal organs (gut, fat body, and salivary glands) are removed to expose the neuromusculature.

Dissected animals may then be fixed for 30 min in freshly made 4% paraformal-dehyde in PBS. Preparations are washed in PBS-TX several times over a period of 1 h and are then incubated for 1 h with the appropriate blocking agent (i.e., PBS-T-BSA). After blocking, the dissected embryos are incubated with the primary antibody (in PBS-T-BSA) for 2 h at room temperature or overnight at 4° C. Fillets are then rinsed four times over a period of 1 h (PBS-T-BSA) and are then incubated with diluted secondary antibodies (in PBS-T-BSA) for 1-2 h at room temperature. The secondary antibody solution is rinsed off (four washes of PBS-T-BSA), and the embryos are mounted and viewed by light or confocal microscopy.

Larval neuromuscular preparations (CNS and body wall) and the stereotypic pattern of innervation of the larval body wall muscles have been described previously (Jan and Jan, 1976a; Johansen et al., 1989a,b). For dissection, the larvae are placed dorsal side up on a 35-mm-diameter petri dish containing a thin layer of Sylgard resin, pinned down at the head and tail using fine insect pins (Fine Science Tools), and cut along the dorsal midline using fine microdissecting scissors. The filleted larva is pinned out flat, and viscera are carefully removed, taking care to leave the CNS intact. In the final preparation, segmentally repeated larval muscles innervated by axons from the CNS are clearly visible (Johansen et al., 1989a,b; Broadie and Bate, 1993). These larvae can then be immunolabeled as described earlier for embryos.

Immunolabeling of the larval and adult CNS often requires sectioning because some antibodies do not readily penetrate these thicker tissues. Paraffin or cryo-sectioning by standard methods are used (see, for example, Buchner et al., in Ashburner, 1989b; Yang et al., 1995), and methods of sectioning and labeling the retina are described in Wolff (2000a,b). A method describing how to make a tool to mount many heads simultaneously, thereby allowing one to section rapidly, can be found in Ashburner (1989b).

Drosophila neurons have been studied by electron microscopy methods when higher resolution is beneficial. For example, fly NMJs have been analyzed by transmission electron microscopy (TEM; Jia et al., 1993; Atwood et al., 1993 and for methods see chapter by Bellen and Budnik, 2000), scanning electron microscopy (SEM; Yoshihara et al., 1997), HRP-enhanced immuno-EM (Lin et al., 1994; Zito et al., 1999), immuno-EM using gold particles (Wan et al., 2000; Pennetta et al., 2002), and immuno-SEM (Broadie and Bate, 1993; Yoshihara et al., 1997). TEM has been used to study the effects of mutations on the steady-state distribution and size of synaptic vesicles in many neuronal mutants. It is useful tool for understanding the role of a protein in the vesicle cycle in conjunction with other assays used to study the presynaptic terminal (i.e., calcium measurements and vesicle dynamics using GFP markers, styryl dyes, and electro-physiology; for methods, see later). Immuno-EM can also be invaluable for determining the subcellular location of a protein; however, one problem is that conditions that preserve tissue morphology the best are frequently too harsh for maintaining optimal antigenicity (see chapter about TEM examination including immuno-EM of Drosophila by McDonald et al., 2000). Immuno-SEM using anti-HRP has been used successfully to outline neurons (Broadie and Bate, 1993; Yoshihara et al., 1997). This method allows one to observe at high resolution the structure and position of dendrites and axons.

C. Transgenic Protein Expression Studies

It is relatively simple to express normal and tagged versions of proteins in the fly nervous system, but it is time-consuming. Genomic and ectopic (heat shock or GAL4/UAS systems) vectors have been utilized extensively for gene expression in Drosophila, and methods of generating transgenic animals are detailed in many chapters (Spradling, 1986; Brand et al., 1994; Phelps and Brand, 1998). Currently, most laboratories use the GAL4/UAS system to create transgenic flies to express proteins in a tissue-specific manner (Brand et al., 1994; Phelps and Brand, 1998). This system is very powerful and easy to use and consists of a yeast transcription factor, GAL4, that can activate transcription in flies (Fischer et al., 1988) but has no endogenous gene targets. Many lines exist in which GAL4 without an enhancer was inserted randomly into the genome by P-element transposition (Spradling, 1986). Lines containing a single insertion have been isolated in which this element is downstream of an enhancer, leading to expression of GAL4 in a temporal and tissue-specific manner depending on when and where the specific enhancer is activated. To utilize this system, another strain of flies is generated with the gene of interest downstream of an upsteam activating sequence (UAS), which allows for transcription of the gene in cells where GAL4 is expressed. Once one has a strain with a gene under the control of UAS, one can cross those flies to flies from a large collection of GAL4 lines. Lines can be obtained that drive expression in a panneuronal manner or in a subset of cells (Brand et al., 1994; Phelps and Brand, 1998; van Roessel and Brand, 2000).

There are several advantages to this system compared to older systems used previously to drive expression by heat shock or genomic promoters. First, one can use the UAS line to drive expression in the cells of choice simply by crossing this line to various GAL4 lines, as opposed to generating multiple transgenic lines each with their own specific promoter. Second, a constitutively expressing genomic construct could potentially be toxic in the wild-type background. However, using the GAL4/UAS system, GAL4 and target gene-bearing flies can be maintained as separate stocks. One point to consider with the GAL4/UAS system is that GAL4 expression is both temperature sensitive and positional sensitive. Because all transgenes, whether driven by a genomic or the GAL4 line, drive expression depending on where they are inserted in the genome, many insertions need to be tested to determine the level of expression induced by the line. The GAL4/UAS system temperature sensitivity can also be taken advantage of as protein expression can be modulated by growing flies at temperature of 18 to 29° C, with 29° C yielding maximum expression. This, for example, can be useful for gain-of-function tests, but may contribute to the variability of expression sometimes seen between flies. A disadvantage of the GAL4/UAS system is that the GAL4 element is inserted in the 5' region of a gene that is expressed exactly where the gene of interest will be driven. These GAL4 insertions could have phenotypes on their own, and in particular, it may be difficult to predict how a GAL4-induced hypomorph may interact with other perturbations in the genome (mutants, overexpression of a protein, etc.). To avoid this problem, one can either generate new lines with an insert of the 5' regulatory region into the construct driving the expression of GAL4. Alternatively, classical methods of using a genomic construct to drive expression have the benefit of possibly driving expression in an otherwise virtually wild-type background.

The transgenic expression of tagged proteins is commonly used to determine the expression pattern and localization of proteins and is useful for the biochemical purification of a protein. The best method for doing this would be to utilize tools of the new gene targeting methods (Rong and Golic, 2000, 2001; Bibikova et al., 2002). Knock-ins of GFP, HA, FLAG, or myo-tagged proteins would generate flies potentially expressing fusion proteins at wild-type levels and with wild-type timing. However, these methods are not yet used routinely because they require more time investment than the generation of lines carrying simple expression constructs. Thus, alternatively one can insert the tagged cDNAs or the genomic region covering the cDNA into vectors with a UAS (UAST) or into vectors requiring the insertion of promoter regions (5UTR) and 3' -untranslated regions (e.g., Casper series, see Flybase for references and sequences of vectors). Tagged versions of proteins driven by these constructs may not necessarily accurately reflect exact wild-type levels, distribution, or timing of expression, but are nonetheless useful. Biochemical experiments, such as immunoprecipitations using commercially available antibodies specific to the tags and gentle elution with the peptides corresponding to the tag, are good methods of identifying interacting proteins (see, for example, Zheng et al., 1995).A complex can then be demonstrated after immunoprecipitation by gel filtration or subcellular fractionation on sucrose and/or glycerol gradients (Zheng et al., 1995, 1998).Specific antibodies are sometimes function blocking, meaning that they could disturb native interactions but tagging a protein could also potentially disrupt function. Elution of coimmunopreciptants from antibodies made against large regions of a protein usually requires harsh treatment, making it more difficult to demonstrate that the interacting proteins are part of a complex. The best approach is to generate specific antibodies and tagged versions of your protein allowing for a confirmation of results by independent methods.

D. Clonal Techniques for Mutant Analyses

Many genes required for neuronal function in Drosophila cause early lethality if they are mutated. One can then either study the embryonic phenotype or use specific methods that allow study of the mutant phenotype in the adult nervous system. Also, if one is interested in studying the role of a protein on the structure or development of neurons in adult tissues, it is difficult because the adult CNS is densely packed with neurons. Silver staining overcomes this issue but can only be used for mutants that allow an adult CNS to develop. Methods addressing both of these issues can be used in flies. To overcome the issue of early lethality, several techniques for generating mutant clones in the nervous system have been described using mitotic recombination (Ashburner, 1989a) via the FLP/FRT system (Golic and Lindquist, 1989). This system leads to expression of the yeast FLP recombinase enzyme that catalyzes the recombination between two 34-bp recognition target sites (FRT sites) (Golic and Lindquist, 1989). Flies are generated with FRT sites on one chromosome and a specific mutation on the homologous chromosome (i.e., flies are heterozygous for the lethal mutation). If recombination is induced after DNA replication, then after mitosis, one sister cell will become homozygous mutant while the other will be wild type. Using this idea, methods of generating eyes composed primarily of mitotic clones of a single genotype have been described, allowing one to study the developmental or physiological consequences of mutation in an adult tissue that would otherwise be lethal (Stowers and Schwarz, 1999; Newsome et al., 2000). The visual system is excellent for the study of neural fate decisions (for review, see Kumar and Moses, 2000) or neuronal pathfinding (for review, see Chiba, 2001). Simple electrophysiological methods also allow one to study both phototransduction and synaptic activity in the retina (see later) (Kelly and Suzuki, 1974; Stowers and Schwarz, 1999). Another recently developed clonal method is mosiac analysis with a repressible cell marker (MARCM; for review, see Lee and Luo, 2001). This method addresses both the issue of early lethality and the one concerning the visualization of individual neurons. This method combines several tools of the fly geneticist, resulting in GFP-marked adult mutant clones in the nervous system. The MARCM system has been used to study autonomous developmental processes in defined neuronal populations in the adult brain, such as axonal pathfinding and dendritic arbor formation (Lee and Luo 1999; Lee et al., 1999, 2000a,b; Scott et al, 2001; Lee and Luo, 2001; Hitier et al., 2001; Jefferis et al., 2001; Grueber et al., 2002; Sweeney et al., 2002).

The MARCM method allows one to analyze the effects of a homozygous mutation in an otherwise wild-type background at a single cell resolution (for review, see Lee and Luo, 2001). Briefly, homozygous mutant clones are generated by mitotic recombination (Ashburner, 1989) via the FLP/FRT system (Golic and Lindquist, 1989). First, a recombinant chromosome is generated during meiosis containing both the recessive mutant gene and a membrane-targeted GFP (mCD8-GFP) on the same arm. The membrane-targeted GFP is a fusion of the mouse T-cell membrane protein, CD8, with eGFP. The expression of mCD8-GFP is under the Gal4/UAS binary system described in Section IV, E (Brand and Perrimon 1993; Brand et al., 1994; Phelps and Brand, 1998; Brand, 1999). Second, two strains of flies are crossed: one Gal4 line, which drives the expression of the yeast transcription factor, and the other driving the expression of the UAS-driven mCD8-GFP fusion protein. This allows one to drive the expression of a transgene by Gal4 in a tissue-specific manner depending on the timing of that promoter. However, in the MARCM system, Gal4 initially cannot lead to transcription of the GFP marker because the other homologous chromosome carries a Gal4 repressor (GAL80) under a ubiquitous driver. This means that all cells that are heterozygous for these two chromosomes will not express the GFP marker. Upon FLP/FRT-mediated mitotic recombination, only one daughter cell receives the repressor (homozygous for repressor) and the other cell is homozygous mutant and expresses mCD8-GFP, but lacks the chromosome containing GAL80. The mCD8-GFP will now be transcribed, translated, and inserted into the membrane outlining the entire cell. FLP expression can be controlled in time and space using a heat shock promoter, which is a promoter that leads to transcription at elevated temperatures. The timing of a particular neuroblast division is known, thereby allowing one to mark only clones actively dividing at the time of FLP expression.

Formation of the Drosophila nervous system occurs by the division of identified neuroblasts that give rise to specific clones of neurons via asymmetric divisions. Each division regenerates a neuroblast and a ganglion mother cell. The ganglion mother divides, yielding two neurons (Goodman and Doe, 1993). Therefore, depending on whether mitotic recombination occurs in a neuroblast, ganglion mother cell, or a dividing ganglion mother cell, a multicellular, two cell, or one cell clone might become GFP marked, respectively. The MARCM system is very powerful but has several drawbacks. If the gene is expressed in a precursor cell, then one has to be aware of the half-life of the protein or mRNA in order to be sure of studying the null rather than a hypomorphic phenotype. By the same token, the mCD8-GFP may not get turned on until the GAL80 level is low enough for GAL4 activation, and then after GFP is expressed, one has to wait for the GFP chromophore to develop (for review, see Tsien, 1998). Despite these slight disadvantages, the MARCM approach has already been used successfully for the study of many genes (Lee and Luo 1999, 2001; Lee et al., 1999; 2000a,b; Scott et al., 2001; Hitier et al., 2001; Jefferis et al., 2001; Grueber et al., 2002; Sweeney et al., 2002) and also for a screen (Lee et al., 2000c).

Was this article helpful?

0 0

Post a comment