Biochemistry of the Drosophila Nervous System

Although genetics is the well-known strength of Drosophila, biochemical studies are feasible in this system and represent a vital research avenue. The quantities of neuronal/brain extract required for many biochemical studies can be readily obtained from flies. Initially, fly head extracts were used to study the biochemical, pharmacological, and electrical properties of various ion channels previously identified and characterized in vertebrate systems (Schmidt-Nielson et al., 1977; Pauron et al., 1987; Pelzer et al., 1989). The idea was to verify the similar nature of the Drosophila proteins and then to identify the Drosophila gene and generate mutants. In other studies, enzyme activity was measured in head extracts to verify the biochemical nature of mutants [e.g., (dnc) Byers et al., 1981; (rdgA) Inoue et al., 1989; (norpA) Inoue et al., 1985; Toyoshima et al., 1990]. More recently, head extracts have been used for many purposes, including identifying proteins playing roles in synaptic transmission (Schulze et al., 1994; van de Goor et al., 1995; Littleton et al., 1998; Phillips et al., 2000), enzyme activity (glutamic acid decarboxylase; Phillips et al., 1993; Featherstone et al., 2000), and RNA metabolism (Zhang et al., 2001). This section describes the method used to generate adult head extracts, first published by the Hall laboratory (Schmidt-Nielson et al., 1977).

Briefly, flies (~1mg/fly) are frozen, and heads (^0.1 mg/head) are collected, homogenized, and then centrifuged as detailed later. This preparation is advantageous because reasonably large quantities of neuronal tissue can be obtained. One disadvantage of this preparation is that a large portion of the material is of retinal origin rather than from other regions of the brain. To circumvent this problem, one may utilize mutants (e.g., eya mutants), which lack compound eyes (although ocelli are present; Stark et al., 1993; Bonini et al., 1998). However, a substantial portion of the fly brain is the optic lobe, and in flies lacking eyes, these neurons are presumably less active, which could be a problem if the biochemical property of interest is regulated by neuronal activity. Another method exists that involves freeze drying using acetone, lyophilization, and the manual separation of eyes and heads (Fujita et al., 1987). This method was used to demonstrate that a substantial portion of the total phosphatidic acid is found in eyes as opposed to the remaining tissue (Fujita et al., 1987). In a later study of rdgA, a mutation in a diacylglycerol kinase, the suprising result was obtained that the substrate (diacylglycerol) levels are normal, while as predicted the level of the product (phosphatidic acid) is reduced. By using the freeze-drying method, the authors demonstrated that this effect of the mutation on phosphatidic acid levels was due to altered phospholipid metabolism in the eye. A take-home message of this study is that the biochemical analysis of flies carrying a mutation in a gene with a predicted function can be beneficial for identifying the true basis of the phenotype (Inoue et al., 1989). This procedure involves freezing flies in a flask with acetone, anhydrous Na2SO4, and liquid nitrogen. This flask is then moved to a freezer (—20 to —25° C) for several days and allowed to dry at room temperature. Under these conditions, the eyes are separated readily from the head, and the quantitative recovery of both proteins and lipids has been demonstrated (Fujita et al., 1987). Many enzymes are not active under these conditions, as the acetone removes the water, yielding most enzymes inactive until one rehydrates the sample. This allows more time for the dissection; however, less material is readily obtained because manual dissection is time-consuming. This method will not necessarily be useful for all procedures but is ideal for lipid biochemistry, especially if the wild-type gene is specifically expressed in either the eye or the brain. In this case, if the substrate of the mutant gene is a molecule(s) that exists and is metabolized in both tissues, then this method allows one to study the effect of the gene on the relevant neuronal tissue in isolation. However, some enzymes do not recover full activity after the acetone treatment and rehydration (Fujita et al., 1987).

A powerful tool for fly neurobiologists is the use of temperature-sensitive paralytics (Suzuki et al., 1971; Grigliatti et al., 1973; Siddiqi and Benzer, 1976). Some examples mutants are in channels upstream of the synaptic cycle or in components of the synaptic vesicle cycle. Some examples include temperature-sensitive alleles of cacophony (a1-calcium channel subunit, von Schilcher, 1976, 1977; Heisenberg and Gotz, 1975; Homyk and Pye, 1989; Smith et al., 1996, 1998; Kawasaki et al., 2002), para (sodium channel subunit; Suzuki et al., 1971), shibire (dynamin; Poodry et al., 1973; Poodry and Edgar, 1979; Koenig and Ikeda, 1989; Poodry, 1990; Kawasaki et al., 2000), comatose (NSF; Ordway et al., 1994; Pallanck et al., 1995; Kawasaki et al., 1998; Littleton et al., 1998), and syntaxin (Littleton et al., 1998). Use of the aforementioned mutants allow one to arrest the synaptic vesicle cycle at various stages (van de Goor et al., 1995; Phillips et al., 2000; Littleton et al., 1998). Using these mutants, flies are shifted to the restrictive temperature (typically 37° C but in some cases 29° C) and fractionated, and the subcellular fractionation pattern the protein of interest is determined. These experiments can be used to begin to decipher at which stage of the vesicle cycle a protein functions (Littleton et al., 1998, 2001). Flies carrying the shibire mutation have been used most frequently in this method. Shibire codes for a temperature-sensitive mutant version of dynamin, a GTPase with a major role in endocytosis (Poodry et al., 1973; Poodry and Edgar, 1979; Koenig and Ikeda, 1989; Poodry, 1990; Kawasaki et al., 2000a,b). When active neurons carrying the shibire mutation are shifted to the restrictive temperature, synaptic vesicles continue to fuse with the plasma membrane but can no longer be endocytosed. This leads to a depletion of synaptic vesicles and accumulation of synaptic vesicle proteins in the plasma membrane. One example of using this tool has been described by Phillips et al. (2000). In this study, the authors sought to determine if stonedB, a protein with a role in vesicle recyling (Stimson et al., 1998; Fergestad et al., 1999), bound to synaptic vesicles before or after endocytosis (Phillips et al. 2000). Previous publications indicated that Stoned was likely to be involved in the synaptic vesicle endocytic pathway and that stonedB is enriched in the P3 fraction, which contains synaptic vesicles (Stimson et al., 1998; Fergestad et al., 1999). The hypothesis was that stonedB associates with synaptic vesicles after endocytosis.

The authors reasoned that if this were the case, stonedB should not be found on vesicles in heat-treated shibire mutants. They took advantage of the use dependency of shibire. Because shibire blocks exocytosis only after endocytosis, shibire only blocks in a use-dependent manner (Salkoff and Kelly, 1978; Salkoff, 1979). If flies are heat shocked in the dark, the visual system will not be depleted of synaptic vesicles, but neurons communicating with other sensory systems will be depleted of synaptic vesicles. However, if stonedB only associates with recycling vesicles, one would expect to see stonedB fail to associate with synaptic vesicles. Contrary to the expectation, in dark-adapted heat shock-treated flies, stonedB remained associated with synaptic vesicles. This study indicated that stonedB associates with synaptic vesicles prior to exocytosis. These studies also indicate that when one uses shibire, one must be aware of the level of synpatic activity in the shibire neuron (Salkoff and Kelly, 1978; Salkoff, 1979; Phillips et al, 2000).

Changes in neuronal activity can alter a biochemical process or protein function. One can purposely alter neuronal activity in flies by using mutants in potassium channels such as double mutants of ether a go-go (eag) and Sh (Shaker), which cause hyperactivity in most of the nervous system (for review, see Wu and Ganetzky, 1992). Reduced activity in select neurons can be achieved using trans-genic flies that express a mutagenized Shaker potassium channel (White et al., 2001a,b; Osterwalder et al., 2001) via an inducible GAL4/UAS system. Other ways of selectively reducing neuronal activity are to express the temperature-sensitive version of shibire (Kitamoto, 2001) or tetanus toxin (Sweeney et al., 1995) under control of the GAL4/UAS system. The benefit of these inducible GLA4/UAS systems is that one can feed the inducing agent to the adult fly, allowing development to advanced stages so that head extracts could be generated.

Flies have been used to study proteins involved in the synaptic vesicle cycle (for reviews, see Littleton and Bellen, 1995; Broadie, 1995; Keshishian et al., 1996; Fernandez-Chacon and Sudhof, 1999; Zhang and Zelhof, 2002) and some of these studies used subcellular fractionation of fly heads (e.g., Schulze et al., 1994; van de Goor et al., 1995; Littleton et al., 1998; Phillips et al., 2000). One such protocol simply uses a glycerol gradient to separate vesicles from the plasma membrane (van de Goor et al., 1995) and is described next.

Method A: Mass Isolation of Fly Heads and Preparation of Head Extract

Approximately 100 mg of heads is generated per gram of flies. Start by expanding flies in bottles or large population cages (Sisson, 2000) to a degree appropriate for the application. Flies are anesthetized (e.g., with CO2) and collected in preweighed 50-ml conical tubes, thereby allowing easy measurement of the weight of the collected flies. The weighed tubes are submerged in liquid nitrogen and stored at —80° C. In a 4° C cold room, each tube is vortexed, causing the heads to detach from the body. The tubes are then emptied onto a stack of prechilled sieves. The stack consists of two brass and stainless-steel sieves with a 710 and 300 ^m mesh, a lid, and a base. Metal sieves can be purchased from companies such as VWR or Fischer [710 ^m, U.S. Standard sieve designation No. 25; 300 ^m, U.S. Standard sieve designation No. 50; brass receiver pan (depth 2 in.)]. Sieves can be prechilled in a —80° C freezer before the experiment is begun or simply cooled by first pouring liquid nitrogen over them. Once all of the flies are on the top sieve, the stack of sieves is shaken vigorously for roughly 30 s, which causes dissociation of the heads from the flies, depositing the heads onto the second screen while the bodies remain on the top screen. The top sieve should be examined for flies with their heads still attached. If necessary, another round of shaking will increase the yield. The frozen heads are then collected into a preweighed tube or homogenizer. The heads can then be either stored at —80° C or homogenized at the desired ratio of buffer to total weight of heads. For example, for total soluble protein extraction, the best yield is obtained at the highest ratio of buffer to total dry weight of heads, but in order to avoid loss of protein activity or nonspecific binding to containers, a ratio of 5 ml buffer/g of heads is a good starting point for satisfactory yields of soluble extract. However, if the first step after the generation of the extract is a method that does not yield concentration and where application of a concentrated sample is critical, such as sucrose or glycerol gradients or gel-filtration chromatography, a lower ratio of buffer to tissue such as 1:1 or 2:1 might be considered.

To begin to generate the head lysate, heads are placed into a mortar prechilled with liquid nitrogen and ground with a frozen pestle until a fine powder is generated. Add more liquid nitrogen as needed. Once a uniform powder is achieved by grinding, weigh the powder in a Dounce homogenizer and add the lysis buffer. Homogenize with the loose-fitting pestle using 5-10 strokes, followed by 10 strokes with the tighter fitting pestle. One can homogenize the heads directly, skipping the mortar and pestle step, but cracking the cuticle using this method is more difficult. Once the lysate is generated, spin for 10min at 1000 x g at 4° C, separating nuclei, cuticle, and unbroken cells from the extract. This post-nuclear extract can then be used for various experiments. If smaller quantities of heads are sufficient, one may simply remove the head from the body by manual dissection using forceps or razor blades under a dissecting microscope. Collect the heads in an Eppendorf tube on ice, and homogenize 10 heads in 100 ^l lysis buffer. For a typical Western, one head per lane on a minigel is a good starting point.

Method B: Subcellular Fractionation of Fly Heads

An initial step at uncovering the function of a protein might be to identify the membrane fraction where the protein is localized. For example, synaptic vesicles and plasma membranes can be separated readily by at least two methods. The first method of fractionating fly heads is by standard methods of differential centrifu-gation. For this method, isolated heads are homogenized at a ratio of 10 ml buffer to 1 g of heads in order to achieve the best separation of membranes and soluble material in the initial steps. Heads (100 mg) are homogenized in 1 ml of buffer [in mM: 10 HEPES, pH 7.4, 1 EGTA, 0.1 MgCl2, 0.1 phenylmethylsulfonyl fluoride

(PMSF)] or in the same buffer except containing Ca+2 instead of EGTA (in mM: 10 HEPES, pH 7.4, 1, CaCl2, 0.1 MgCl2, 0.1 PMSF). A comparison of the distribution of the protein in these two buffers will test for the effects of calcium on the subcellular compartmentalization of the protein of interest. This test might be used if it is supected that the protein may function in calcium-dependent events such as synaptic activity or cell adhesion. Alternatively, if the effects of calcium are not to be tested, magnesium can be left out of the buffer, and if a higher ionic strength buffer is desired, 150 mM NaCl or K glutamate may be added. Higher ionic strength buffers can be used, for example, if one intends to use various fractions for immunoprecipitations; however, if one intends on using fractions for an ion-exchange column, for example, leaving out the salts is desirable. Homogen-ates are then centrifuged at 1000 x g to produce the P1 pellet fraction. The supernatant (S1) is centrifuged at 25,000 x g at 4° C for 40min to produce the P2 (heavy membrane) pellet fraction. The resulting S2 supernatant is then centri-fuged at 4° C at 125,000 x gave for 1 h to obtain the P3 (light membrane and synaptic vesicle) pellet and the final supernatant, S3 (soluble fraction).

If the distribution of all proteins and membrane relative to each other is desired, then head extracts can be fractionated on a density gradient. Powdered heads (1 g) can be resuspended in 1 ml of lysis buffer (in mM: 150 NaCl, 10 HEPES, ph 7.4, 1 EGTA) and homgenized 5-10 strokes using the loose Dounce pestle, followed by 10 strokes with the tight-fitting pestle. The postnuclear supernatant (10min at 1000 x g) is layered carefully onto a 5-25% glycerol gradient made in lysis buffer, overlaying a 50% sucrose cushion also made in lysis buffer. For optimal separation, use a volume of supernatant corresponding to less than 5-10% of the volume of the gradient and a cushion representing 10% of the gradient volume. This gradient can be made most simply in the following manner (e.g., for a Beckman SW41 rotor): Place 1 ml 50% sucrose in lysis buffer at the bottom of the gradient. Make 5 and 25% glycerol solutions in lysis buffer. Make a 1:1 mixture of the 5 and 25% stocks and then a 1:1 mixture of this solution and either the 5 or 25% stocks. Layer 2.1 ml of each of these five solutions sequentially on top of the cushion, beginning with the 25% glycerol solution and ending with the 5% glycerol solution. Be careful to add these solutions slowly to avoid mixing. After pouring, allow the gradient and cushion to sit in the cold room for at least 2 h prior to adding the supernatant. Add 0.5 ml of the postnuclear fraction on the top of the gradient and spin for 2h in a swinging bucket rotor at 160,000 x gave (36,000 rpm in a Beckman SW41Ti). Fractionate the gradient by removing 0.5-ml fractions, and assay each by Western analysis. Alternatively, if a tabletop ultracentrifuge is available, this separation using glycerol gradients is more rapid. For example, the gradients can be spun at 50,000 rpm in a TLS-55 rotor using a Beckman tabletop ultracentrifuge for 30min. Fractions containing synaptic vesicles are identified using antibodies specific for cysteine string protein or synapto-brevin, whereas fractions containing plasma membrane are identified using antibodies against syntaxin, HRP, or the Na/K ATPase (see Table II). The

Table II

List of Antibodies to Drosophila Proteins Found in Neurons and Glia



Expression pattern



Mouse mAB

Nuclei of four neuroblasts

Skeath and Carroll (1992)


Rabbit pAB


Gonzalez-Gaitan and Jackle (1997)

a PS1

Mouse mAB DK1A4

NMJ postsynaptic

Brower et al. (1984); Beumer et al. (1999)

a PS2 integrin

Rat mAB PS2hc2

NMJ postsynaptic

Bogaert et al. (1987); Beumer et al. (1999)

a spectrin

Mouse mAB

NMJ, pre- and postsynaptic

Dubreuil et al. (1997); Featherstone et al. (2001)


Rabbit pAB

Two neurons around mushroom body

Waddell et al. (2000)


Mouse pAB

Embryonic nerve tracts

Bouley et al. (2000)

Was this article helpful?

0 0

Post a comment