Defective Glia in the Drosophila Brain Degeneration Mutant drop-dead

Defective Glia in the Drosophila Brain Degeneration Mutant drop-dead

Neuron,Vol. 10, 839-850,May,1993,Copyright© 1993by Cell Press Defective Glia in the Drosophila Brain Degeneration Mutant drop-dead Robert L. Buchanan...

14MB Sizes 0 Downloads 5 Views

Neuron,Vol. 10, 839-850,May,1993,Copyright© 1993by Cell Press

Defective Glia in the Drosophila Brain Degeneration Mutant drop-dead Robert L. Buchanan and Seymour Benzer California Institute of Technology Division of Biology Pasadena, California 91125

Summary To understand better the cellular basis of late-onset neuronal degeneration, we have examined the brain of the drop-deadmutant of Drosophila. This mutant carries an X-chromosomal recessive mutation that causes severe behavioral defects and brain degeneration, manifested a few days after emergence of the adult. Analysis of genetically mosaic flies has indicated that the focus of the drop-dead mutant phenotype is in the brain and that the gene product is non-cell autonomous. We examined the adult drop-dead mutant brain prior to onset of symptoms and found that many glial cells have stunted processes, whereas neuronal morphology is essentially normal. Adult mutant glial cells resemble immature gila found at an earlier stage of normal brain development. These observations suggestthat defective glia in the dropdead brain may disrupt adult nervous system function, contributing to progressive brain degeneration and death. The normal drop-deadgene product may prevent brain degeneration by providing a necessary glial function. Introduction Neurodegeneration is a characteristic of numerous inherited, late-onset CNS disorders in organisms as diverse as humans (McKusick, 1990) and nematodes (Driscoll and Chalfie, 1992). Several Drosophila mutants exhibit late-onset and tissue-specific neurodegeneration (Hotta and Benzer, 1970, 1972; Benzer, 1971; Heisenberg, 1980; Coombe and Heisenberg, 1986; Inoue et al., 1989; Steele and O'Tousa, 1990). The molecular basis for some of these disorders is beginning to be understood, in part through the use of genetic techniques that allow the rapid identification and cloning of the genes involved. In Drosophila, for example, genes have been characterized that play important roles in photoreceptor degeneration (Bloomquist et al., 1988; Zuker et al., 1985; Montell and Rubin, 1989; Vihtelic et al., 1991; Freeman et al., 1992). Structural defects in rhodopsin lead to photoreceptor degeneration both in Drosophila (O'Tousa et al., 1989) and in humans (Dryja et al., 1990). This suggests that mechanisms controlling neurodegeneration may be shared between Drosophila and higher vertebrates. The Drosophila drop-dead mutant provides a convenient genetic model system in which to study lateonset brain degeneration. Newly emerged adult mutant flies exhibit normal motor behaviors such as flying, phototaxis, sexual courtship, and mating. After

some days, however, the mutants begin to show deficiencies in flight as well as staggering locomotor patterns, accompanied by the appearance of brain lesions and rapid death. Whereas the onset of symptoms in a mutant population is stochastic, the gene is fully penetrant; within several days, all mutant flies are dead (Benzer, 1971; Hotta and Benzer, 1972). Analysis of genetically constructed mosaic flies has suggested that the focus of the drop-dead syndrome maps to the brain, i.e., a mutant brain in an otherwise normal bodywill produce brain degeneration and death, and bilateral mutant brain tissue must be present for these defects to occur. These observations have been interpreted as suggesting that normal brain tissue may provide a factor that prevents degeneration elsewhere in a mosaic brain (Hotta and Benzer, 1972). Although neurodegeneration is most often associated with defective neuronal function, abnormal glial cell function has been implicated in an increasing number of vertebrate neuropathies (Papp et al., 1989; Siman et al., 1989; Knapp et al., 1990). In both vertebrate and invertebrate nervous systems, glia normally provide structural and metabolic support for neurons, regulate nervous system homeostasis, and respond to several neuromodulators (Kuffler, 1967; Tsacopoulus et al., 1988; Pentreath, 1989; Carlson and St. Marie, 1990; Barres, 1991). Although the functions of adult Drosophila glia have not been well studied, there have been extensive analyses of developing insect glial cells by anatomical (Nordlander and Edwards, 1969; Grenningloh et al., 1990), immunological (Grenningloh et al., 1990; Meyer et al., 1987), and molecular genetic (see Grenningloh et al., 1990; Nambu et al., 1990; Rothberg et al., 1990) approaches. These studies have revealed several types of insect glia that have multiple roles in nervous system development. Glia may provide critical functions in invertebrate nervous system maintenance and repair, since they proliferate in response to damaged tissue (Carlson and St. Marie, 1990) and may transport proteins to injured neurons (Griffiths, 1979). It is therefore plausible that defective glial cell function could perturb normal nervous system development and the maintenance of mature neuron function. We have examined the morphology of the dropdead adult brain and found structural defects in glial cells, prior to the onset of neuronal degeneration within the adult brain. These and other data suggest that the drop-deadgene is necessary for normal brain function and that abnormal drop-dead glia may be responsible for the progressive brain deterioration and death in the adult mutant. Results Onset of Lethality in drop-deadMutant Adults Wild-type Drosophila will normally live for several

Neuron 84O



...... Z01 0.5 In(l) drd xl 7 drop.dead --


drd ~

J mutants

c o u. 0.1













10 12



14 16


Time post-ecloslon(days)

Figure 1. Survival Curves of the drop-dead Mutants drd~ and In(1)drdX~Compared with Wild Type weeks in the laboratory, but flies bearing the dropdead mutant alleles drd 7 or In(1)drd x~ exhibit a short lifespan; most die within the first week of adulthood (Figure 1). Newly ecIosed drop-dead mutant flies behave normally, but with time, exhibit behavioral defects such as sluggish movement or inability to walk or fly. Death typically occurs within a few hours after initial manifestation of these defects. Within either mutant population, the onset of symptoms is stochastic, in agreement with previous observations (Benzer, 1971; Hotta and Benzer, 1972). To investigate whether the drop-dead mutants are defective in normal development before adulthood, drop-dead mutant flies were examined for pre-adult lethality or a delay in the time required to progress through the embryonic, larval, and pupal stages (see Experimental Procedures). No pre-adult lethality or gross delay in normal development was detected in drop-dead mutant flies bearing either mutant allele. These results are consistent with the absence of such deficits in the drop-dead allele drd 2 (listed as drd 7°7 in Homyk et al., 1986, and drd 2 in Lindsley and Zinn, 1992). Nevertheless, we observed previously unreported features of the drop-dead mutant phenotype, including a slightly bent wing margin, frequent hypertrophyofthe abdomen, reduced body size, and recessive sterility in homozygous mutant females.

Abnormal Brain Morphology in the Aging drol~lead Adult Brain The adult fly brain and optic lobes consist of outer cortices containing neuronal cell bodies and inner synaptic neuropils (Figure 2A; Strausfeld, 1976). The optic neuropils are involved in visual processing (Strausfeld, 1976; Kankel et al., 1980), whereas the cen-

tral brain neuropil contains important association centers. Glial cell perikarya are found particularly at boundaries between the brain cortex and neuropil, within the brain cortex, and within the perineurium (Strausfeld, 1976). The perineurium surrounds the brain cortex and has been compared functionally to the vertebrate blood-brain barrier (Carlson and St. Marie, 1990). A number of distinct glial cell types that extend complex cell ramifications to multiple neurons have been characterized in various insect species (Meyer et al., 1987; Strausfeld, 1976; Hoyle, 1986; Boschek, 1971; Carlson and St. Marie, 1990). Mutant brains were surgically excised from l-weekold drop-dead males that had begun to exhibit locomotor or flight defects. These were sectioned and stained with toluidine blue for microscopic examination. Tissue from such mutant flies exhibited brain abnormalities of variable severity, such as more intense staining of the cortex (Figure 2B), intense staining of both the cortex and neuropil (Figure 2C), or highly intense staining accompanied by numerous brain lesions (Figure 2D). Typically, lesions were most abundant in the central brain, although the optic lobes also exhibited degeneration. The unequal distribution of lesions suggests that brain tissues may differ in susceptibility to defective drop-deadgene function, or that degeneration begins in a relatively central region and spreads outward. Visualization of Glia in Newly Eclosed Flies with Toluidine Blue Consistent with earlier observations of the absence of gross brain pathology in newly eclosed mutant flies (Hotta and Benzer, 1972; Benzer, 1971), the brains of newly emerged mutant flies did not exhibit the tissue degeneration associated with older mutant adult brains (Figures 2B-2D). In wild-type brain sections, darkly staining cells were prominent in the neuropil and at the cortex-neuropil border (Figure 3A). Intensely stained cells could also be identified in the brain cortex and perineuriurn. In contrast, toluidine blue-stained brain tissue from newlyemerged mutant flies exhibited weaker staining intensity of the corresponding cells (Figure 3). The abnormal staining intensity of these brain cells varied among individual mutant flies. Based on the strong affinity of normal insect glial cells for thiazine dyes (Trujillo-Cen6z, 1965; Meyer et al., 1987; Tolbert and Oland, 1990), previous anatom ical descriptions of insect gila (see Meyer et al., 1987; Strausfeld, 1976; Fischbach and Technau, 1984; Miller, 1965), and our own electron microscopic studies, these cells were tentatively identified as gila. Optic lobe gila have been anatomically studied in other flies (Trujil Io-Ce n 6z, 1965; Bosch e k, 1971; St ra u sfe Id, 1976). These cells did not exhibit reproducible abnormal staining intensity in newly emerged mutant flies. These observations suggest that the drop-dead mutation alters the normal staining properties of brain gila and that this defect is manifest prior to onset of the degeneration that occurs in older mutant adults.

Brain Degeneration Mutant 841

Figure 2. Adult Brain Degeneration in the Aging drop-dead Mutant Horizontal head sections of 1-week-old males stained with toluidine blue (see Experimental Procedures). (A) Wild type. (B) drop-dead (drd T) mutant. Note increased intensity of staining in the cortex. This male had begun to exhibit locomotor abnormalities. (C) Another mutant individual (drdl). Note increased staining in the cortex and neuropil. (D) This mutant (drd 7) individual, although still alive, had advanced brain degeneration, as evidenced by intense tissue staining and vacuoles within the brain. C, brain cortex; NP, brain neuropil; R, retina. Bar, 50 p.m.

Individual Glia Visualized by Golgi Impregnation The Golgi t e c h n i q u e randomly impregnates individual neurons and gila w i t h a silver chromate precipitate, rendering an entire cell visible to light microscopy (for review see Strausfeld, 1980). Since the

pioneering w o r k on insect neurons by Ramon y Cajal (Ramon y Cajal and Sanchez, 1915), the technique has been used to characterize neuron morphology in Drosophila (Fischbach and Dittrich, 1989; Coombe and Heisenberg, 1986), as well as in other insects and verte-

Neuron 842

BrainDegenerationMutant 843

brates (Strausfeld, 1976, 1980; Tolbert and Oland, 1990). Since adult brain glial cells have not been well characterized in Drosophila, we used the Golgi technique to visualize their morphology in normal and drop-dead mutant flies, using the method of Fischbach and Dittrich (1989). Impregnation of wild-type brain tissue often stained large numbers of brain gila, which formed a dense network in the brain cortex and between the cortex and neuropil. Figure 4A shows an individual Golgi-impregnated brain cortex glial cell from a wild-type brain. The cell exhibits a dense cell nucleus and diaphanous processes that envelop neighboring neuronal cell bodies. In drop-dead mutant brain tissue, although many Golgi impregnated brain gila had apparently normal cell morphology, abnormal cortex gila were also present (Figures 4B4D). These mutant gila exhibited abnormally dense impregnation of their cell processes (Figure 4B), lacked the fine structure (Figure 4C) associated with normally impregnated glial cells (Figure 4A), or exhibited stunted cell processes (Figure 4D). In contrast with the gila, neurons did not show any qualitative differences in impregnation, in over 50 newly eclosed wild-type and mutant brains examined. Glial Cell Ultrastructure in the Newly Eclosed Adult drop-dead Brain To characterize further abnormal drop-dead glial cell morphology and to quantitate abnormal cortex gila, wild-type and drop-dead mutant brains were analyzed by transmission electron microscopy. Brain glial cells are osmiophilic and readily identified by their greater electron density than surrounding neurons. In the brain cortices of wild-type flies, neuronal cell bodies were seen to be completely enveloped by electrondense glial cell processes (Figure 5A). These cells exhibited a highly extended cell structure corresponding to that observed with the Golgi technique. The electron-dense glial cell nucleus, extensive array of electron-dense cytoplasmic processes, and close apposition to adjacent neuronal tissue (Figure 5; Figure 6) are consistent with other observations in insects (Boschek, 1971; Carlson and St. Marie, 1990; Fischbach and Technau, 1984; Hoyle, 1986; Lane, 1985; Strausreid, 1976; Tolbert and Oland, 1990; Trujillo-Cen6z, 1965,1985). By semiserial electron microscopy, we determined that approximately 1%-5% of the brain cells in wild-type and mutant tissue are cortical gila. In newly emerged drop-dead mutant flies, the electron

microscope revealed that over 70% of the brain cortex glial cells examined exhibited abnormal ultrastructure (Figures SB-SD) days before the onset of brain degeneration in older adults (Figure 2). Glial cellswere found with stunted processes that failed to wrap neighboring neuronal cell bodies completely (Figures 5B and 5C), or exhibited more severe morphological defects, such as condensed cytoplasm, vacuoles, abnormal cytoplasmic bodies (myelin figures), and pycnotic nuclei (Figures SC and SD). The morphology of these abnormal cortex gila resembles descriptions of apoptotic cell death in other systems (Duvall and Wyllie, 1986). In Figure 6B, the tip of a mutant cortex glial cell process terminates between two neuronal cell bodies, leaving an extracellular space of abnormally grainy and less electron-dense extracellular matrix (compare Figures 6A and 6B). In contrast with the gila, neurons in the newly eclosed drop-dead fly had normal morphology, although neurons occasionally contained abnormal cytoplasmic multilamellar bodies (Figure 6C). Semiserial electron microscopy of dropdead mutant brains indicated that such bodies were not restricted to any particular area of the brain. Ultrastructure in the Developing Brain To determine whether defects in ultrastructure are manifest at earlier stages of mutant brain development, glial and neuronal cell morphologywere examined in the late pupa. In the wild-type pupal brain, 70-80 hr after puparium formation, about half of the glial cells at the cortex-neuropil border were found to invest adjacent neuronal tissues with highly extended cell processes. For example, in Figure 7A, a wild-type pupal glial cell atthe cortex-neu ropil boundaryexhibited features in common with adult brain gila, namely electron-dense cytoplasm and an extensive array of cell processes. This wild-type pupal glial cell clearly separates the immature brain cortex from the adjacent neuropil, in addition to enveloping a neighboring axon bundle. However, at this same stage of development, about half of the wild-type gila exhibited short, less elaborate cell processes. In Figure 7B, for example, the pupal glial cell does not provide a complete barrier between the brain cortex and the neuropil, nor do the glial cell processes surround neurons as observed later in the adult brain. These are presumably immature glial cells. Examination of several 70-80 hr pupal brains from both of the drop-dead mutant alleles failed to reveal any difference from wild

Figure 3. Abnormal Glial Cell Staining in the Newly Eclosed drop-deadMutant Brain Transverse head sections of newly eclosed adult males stained with toluidine blue. (A) The wild-type brain contains darkly staining cells at the cortex-neuropil border (arrow) and within the neuropil. These are identified as gila (see text). (B) In drop-deadmutant (In(7)drdxl) brain, gila stain less intensely. (Note that toluidine blue staining in the mutant becomes more intense as the brain ages and degenerates,which does not occur in wild type [Figure 2].) C, cortex; E, esophagus; NP, brain neuropil. Bar, 25 ~.m. Figure 4. Golgi-lmpregnated Gila in Newly Eclosed Mutant Brain Cortex Exhibit Abnormal Cell Structure (A) Wild-type glial cell. Note the densely impregnated cell nucleusand array of thin, relatively transparent cell processesthat surround several unstained neurons. (B and C) Three glial cells in drop-dead(drd~mutants. Note that some cell processesare more densely impregnated in (B). In (C), the glial cell processarray is dense and disorganized. In (D), a mutant cortical glial cell revealsonly a single process. Many other brain gila in the mutant impregnate normally (seetext). Bar, 10 I~m.

Neuron 844

t \

G "--7


Figure 5. Abnormal Glial Cell Ultrastructurein the Newly Ecloseddrop-dead Brain (A) A wild-type glial cell of the brain cortex is electron dense and exhibits extensivecytoplasmicprocessesthat envelop neighboring neuronal cell bodies (arrows). (B) drop-dead (drd~)mutant brain cortex.The glial cell has abnormallyshort processes(arrows).(C and D) drop-dead(In(1)drdxT)mutant glial cellsdisplayshort cell processes(arrows)and signsof advanceddegeneration,such as vacuolated cytoplasm (C) and abnormal cytoplasmic bodies ([D], arrowhead), indicated by arrows. G, glial cell. Bar, 0.5 p.m.

type. In both wild-type and mutant tissue, unidentified dead cells that exhibited highly condensed, electron-dense ultrastructure were occasionally encountered. These dead cells reflect programmed cell death in the brain, a normal process associated with nervous system development in Drosophila (Fischbach and Technau, 1984; Hofbauer and Campos-Ortega, 1990). Examination of brain cell morphology prior to the pupal stage (late third larval instar) also failed to identify morphological differences between drop-dead and wild-type brain tissue.

Discussion Flies bearing the X-chromosome-linked drop-dead mutation exhibit late-onset brain degeneration during the first week of adulthood. Using two different techniques, staining with toluidine blue or Golgi impregnation, abnormal glia were detected in newly eclosed drop-dead mutant flies, prior to the onset of brain degeneration. Ultrastructural analysis of the mutant brain revealed glia having abnormally short cell processes or exhibiting characteristics of ad-

Brain DegenerationMutant 845

Figure 6. Short Glial Cell Processesand Abnormal ExtracellularMaterial in the Newly Ecloseddrop-dead Brain (A) Wild type. The space between neuronal cell bodies is filled with electron-denseglial cell processesof variablethickness (arrows). (B) drop-dead (In(7)drdxT) mutant. Note the truncated glial process (arrow). The space between neuronal cell bodies contains a low density extracellularmatrix (arrowheads)that is not observed in wild type. (C) drop-dead (drd~)mutant. Neuronal cell morphology is normal at a gross level,although some mutant neuronal cell bodies contain abnormal cytoplasmicinclusions (myelinfigures)that are not found in wild-type neurons, m, myelin figure; N, neuronal cell body. Bar, 0.1 I~m. vanced cell degeneration. Earlier in development, in the 70-80 hr pupal brain, drop-dead and wild-type brain tissue had similar ultrastructure, suggesting that glial cell defects in the mutant become manifest between the late pupal stage and the first several hours of adulthood. These observations indicate that glial cell morphology is defective in the newly emerged drop-dead adult brain, before the onset of neurodegeneration and subsequent death. The identification of defective brain glial cells is consistent with previous genetic data identifying the brain as the focus of the

drop-dead mutant phenotype (Hotta and Benzer, 1972). The aging drop-dead mutant exhibited abnormal patterns of toluidine blue staining in the brain. In the newlyeclosed mutant, brain glia stained less intensely than normal, whereas neurons did not (Figure 3). With age, the mutant brain developed increasing intensity of staining, becoming much more intense than agematched normal flies (Figure 2). The reasons for these changes are presently unknown. The initially diminished staining of glia could be because of a lack of

Neuron 846

Figure 7. Glia with Short Processes Are Found in the Developing Wild-Type Brain (A) 70-80 hr after puparium formation, this wild-type glial cell has highly extended processes (arrows) that separate the neuronal cortex from the inner synaptic neuropil. This is typical in mature adult glia. (B) The processes in this wild-type pupal glial cell (arrows) are much shorter, suggesting that it is immatu re. Note the absence of glial cell processes between adjacent brain cortical cells (arrowheads), as is often seen in the adult drop-dead mutant brain (Figure 6). Axb, axon bundle; C, cortex; G, glial cell; NP, brain neuropil. Bar, 0.5 pm.

some cellular c o m p o n e n t or a deficiency in glial cell processes. The later, increasingly intense staining of the entire mutant brain could be the result of the accumulation of degradation products w i t h i n cells. The three histological techniques used, t o l u i d i n e blue staining, Golgi impregnation, and electron microscopy, varied in their ability to detect abnormal brain glia in the drop-dead mutant. Toluidine blue staining of newly eclosed mutant brain tissue indicated that cortex glia, cortex-neuropil border glia, neurOpil, and perineurium glia stained abnormally w h e n compared w i t h wild-type tissue. Electron microscopy revealed that 70% of brain cortex glia exhibited abnormally short processes and/or condensed cell cytoplasm, cytoplasmic vacuoles, and pycnotic



nuclei. Electron microscopy also indicated that cortex-neuropil and perineurium brain glia in the mutant were often more compact than normal (data not shown). Compared with toluidine blue staining or electron microscopy, the Golgi t e c h n i q u e identified fewer abnormal glial cells. This may reflect the inability of the Golgi technique to impregnate successfully some types of degenerating insect cells (Strausfeld, 1980; Coombe and Heisenberg, 1986). Optic lobe glia in diptera have been described in detail (Boschek, 1971; Carlson and St. Marie, 1990). We did not detect reproducible abnormalities in these cells b y t o l u i d i n e blue staining or electron microscopy. Taken together, these data indicate that various types of brain glia are abnormal in the drop-dead mutant, w i t h brain cortex






Figure8. Adult drop-dead Mutant Glial Cells Resemble Immature Gila Found in the Pupal Brain Schematic tracings of electron micrographs of wild-type and drop-deadmutant gila from the late pupal and adult stages. In the 70-80 hr pupa, approximately half of the gila in both wild type and mutant exhibit a compact shape, with short cell processes (A). In the wild-type adult, the brain cortex gila develop extensive cell processes (B). In approximately 70%; of the adult drop-deadmutant brain cortex glia, these are lacking (C). Some adult dropdead mutant brain cortex gila also exhibit ultrastructural features consistent with apoptotic cell death (see text). Dots indicate glial cell processes going out of view. N, glial cell nucleus.

Brain DegenerationMutant 847

glia being the most structurally defective. Although these glia represent a small fraction (1%-5%) of the total population of brain cells (neurons plus glia), Golgi impregnation of single cortex glial cells indicates that each may surround many neurons (Figure 4A). Therefore, abnormalities in brain cortex glia, as observed in the drop-dead mutant, have the potential to adversely affect the function of large numbers of brain neurons. During normal nervous system development, immature insect glia tend to have relatively short cell processes, which branch out as they become mature adult cells. For example, in the Calliphora pupal nervous system, there are short glial processes that do not form extensive connections with neurons or other glia until later in development (for review see Lane, 1985). In the antennal lobe of Manduca, glia extend cell processes to envelop olfactory glomeruli as they mature (for review see Tolbert and Oland, 1990). In both these systems, normal CNS development appears to require extensive changes in glial cell morphology and presumably function. In agreement with these studies, 70-80 hr Drosophila pupal brain glia in both wild type and mutant exhibited shorter cell processes than found later in the normal adult brain. The observation that short glial cell processes occur in both immature wild-type pupal glia and in adult drop-dead brain glia implies that glial cell development in the mutant may be arrested (see Figure 8). In this model, drop-deadgene function would be required to support normal glial cell maturation. In the mutant brain, immature and defective glia would not support nervous system function, leading to brain degeneration and death. Alternatively, the drop-dead mutant defect could cause retraction of normally formed glial cell processes during early adulthood. Defective maturation has been proposed to explain neurological defects in the jimpy mouse mutant. In the jimpy mouse, defective oligodendrocyte development gives rise to an adult CNS containing immature and defective glia that are unable to interact normally with CNS neurons (Knapp et al., 1990). Despite the observation that morphological defects in mutant brain glia are detectable prior to the onset of brain degeneration, ultrastructural studies alone cannot determine whether defects in glia, neurons, or both cell types are the primary causes of brain degeneration. For example, vertebrate neurons have the capacity to regulate glial cell development and function (see Gasser and Hatten, 1990; Barres, 1991). If Drosophila neurons have similar properties, the glial abnormality could result from a neuronal defect. Alternatively, both neurons and glia may require dropdead gene function, but glia could be more sensitive to abnormal drop-dead gene product activity. Although neuronal cell morphology in the newly eclosed mutant brain is grossly normal compared with cortex glia, neurons containing abnormal cytoplasmic bodies (myelin figures) were occasionally observed at eclosion. Such abnormal cell bodies have been described in traumatized insect neurons, and these may reflect

early neuron autolysis prior to cell death (Stocker et al., 1978). Diffusible and/or secreted proteins are known to be important for the differentiation of neurons, glia, and other cell types in Drosophila (see Grenningloh et al., 1990; Freeman et al., 1992; Padgett et al., 1987; Rothberg et al., 1990) and vertebrates (see Anderson, 1989). In brains genetically mosaic for the drop-dead mutation, wild-type tissue can rescue the mutant portion of the brain from degeneration, indicating that the effect of the drop-dead gene product is not cell autonomous (Benzer, 1971; Hotta and Benzer, 1972). Wild-type tissue could provide a diffusible, brainspecific factor encoded by the drop-dead gene, which is necessary for normal development and/or maintenance of brain glia, thus preventing degeneration in mosaic brains. Such a putative factor might be synthesized by neurons, glia, or both cell types. An alternative model that could explain the ability of wild-type tissue to prevent degeneration in a mosaic brain is cell migration. Normal cells might "seed" the drop-dead portion of the brain with normal glia during development. Although glial cell precursors are thought to undergo migration during Drosophila em bryogenesis (Jacobs et al., 1989), analysis of adult mosaic brains containing internally marked tissue (Kankel and Hall, 1976) does not support the concept of very extensive cell migration; fairly distinct mosaic boundaries separating marked and unmarked tissue are usually observed. Nevertheless, in response to defective neurons, insect glia are known to express new functions (Griffiths, 1979; Carlson and St. Marie, 1990), and it is conceivable that migration of normal cells could be stimulated in response to drop-dead tissue. Such an analysis will require appropriate internal brain cell markers to distinguish between mutant and normal neurons and glia in mosaic brains. Recent molecular cloning of the drop-dead gene (Buchanan and Benzer, unpublished data) allows us to ask what cells express drop-dead mRNA and to address whether the drdl and In(1)drdx~alleles are nulls (loss of function alleles). Northern blot hybridization using a drop-dead cDNA as probe did not detect any transcript in In(1)drdx~flies. In the other allele, drd ~, a transcript is detected. However, it is presumably not functional, since the drd 7 and In(1)drd~7 mutant phenotypes are closely similar. It is likely that both drop-dead alleles are nulls or severe hypomorphs. The drop-dead mutation is one of several mutations that are known to be important for adult survival (Flanagan, 1977; Homyk et al., 1986; Henkemeyer et al., 1987; Carthew and Rubin, 1990; Freeman et al., 1992; Tei et al., 1992), or to maintain brain integrity (Heisenberg, 1980; Coombe and Heisenberg, 1986) in Drosophila. We have provided morphological evidence suggesting that late-onset brain degeneration in the drop-dead mutant may be triggered by defects in glia rather than neurons. It would seem certain that an array of cell factors is important for the maintenance of adult nervous system function and that the dropdead sort of phenotype may be used to screen for

Neuron 848

t h e c o r r e s p o n d i n g g e n e s . Recent e x p e r i e n c e w o u l d i n d i c a t e t h a t m a n y o f t h e s e D r o s o p h i l a g e n e s m a y be expected to have vertebrate homologs.

Experimental Procedures Drosophila Stocks and Genetics The wild-type (Canton-S), drop-dead (drY), and FM7c fly strains are described in Lindsley and Zimm (1992). The drop-dead mutant allele In(1)drd xl was isolated in this laboratory using a standard F1X-ray screen for adult lethal mutants (Grigliatti, 1986) that did not genetically complement drd 7.The In(1)drd x~defect maps to the same chromosomal region as drd T(13A-B), as determined by deficiency mapping using flies bearing X-Y translocations (Stewart and Merriam, 1975). Mutant stocks were genetically balanced over the FM7c X-chromosomal balancer (drdVFM7c and In(1)drd~VFM7c) and reared on cornmeal-yeast-agar medium at 25°C.

Analysis of Pre-Adult and Adult Lethality Approximately 100 newly eclosed male flies were collected from uncrowded cultures and placed in food vials (20 flies per vial). The number of dead flies was counted each day. Each drop-dead allele was assayed for pre-adult lethality by counting the total number of eclosed drop-dead males compared with the total number of eclosed FM7c balancer (drd+) males. No evidence for pre-adult lethality was found in either mutant allele. The approximate time required to progress through embryogenesis and reach the third instar larval stage was evaluated for d r d 7 and In(1)drd x~ male flies. Wild-type or FM7c balanced drop-dead females (10-15 total) were allowed to mate and lay eggs for 1 day, and the numbers of male third instar larval progeny were scored 4-5 days later. Male larvae carrying the FM7c chromosome were genetically white and could be scored by the loss of yellow pigment in the larval Malpighian tubules. No significant differences were found in the numbers of wild-type, drop-dead, or FM7c third instar larbae. The length of the pupal period was evaluated by placing late third instar (white +) male larvae from Canton-S, drd~lFM7c, or In(1)drdxTIFM7c stocks into separate vials with food. In all three stocks, the length of the pupal period (from pupariation to emergence) was approximately 105-115 h r at 25°C, in excellent agreement with literatu re values for wild type (Ashburner, 1989).

Tissue Preparation for Light and Electron Microscopy Adult; pupal, and third instar larval tissues were prepared for light and electron microscopy as described previously (Renfranz and Benzer, 1989), except that pupal head s were dissected under freshly prepared 4% paraformaldehyde plus 2% glutaraldehyde. For light microscopy, 1 p.m sections were cut from several wildtype or mutant brain regions and stained in parallel for approximately 1 rain on a warm plate with an aqueous solution of 1% toluidine blue plus 1% Borax. Ultrathin Epon plastic sections were poststained in 2% uranyl acetate, followed by Reynolds' lead citrate (Reynolds, 1963), and stabilized for transmission electron microscopy by brief carbon coating. Tissue specimens were examined on a Philips 201C electron microscope at 40-80 kV. Sections from several brain regions were examined from 10 drd ~ and 10 In(1)drd~7 male flies and 4 wild-type flies. In addition, semiserial brain sections were prepared at 10-15 p.m intervals from 3 Canton-S, 3 drd ~, and 2 In(1)drd ~ male flies. Glial cell bodies in the brain could be clearly identified by their electrondense cytoplasm and characteristic cell shapes (Carlson and St. Marie, 1990; Boschek, 1971; Hoyle, 1986).

Golgi Impregnation of Brain Tissue Golgi impregnated brain tissue was prepared according to a modification (Fischbach and Dittrich, 1989) of the Golgi-Colonnier method (see Strausfeld, 1980), except that after the second cycle of impregnation, fly heads were dehydrated and embedded in Epon plastic (Renfranz and Benzer, 1989). Approximately 50 successfully impregnated wild-type, drd 1,and In(1)drd x~brains w e r e sectioned at 5 I~m and visualized by Nomarski optics.

Acknowledgments We gratefully acknowledge the expert technical assistance of Eveline Eichenberger, Pat Koen, and Rosalind Young, as well as many helpful discussions with our research group. We also thank Edward Lewis for the use of his X-ray facility and important discussions, Karl Fischbach for a detailed Golgi impregnation protocol, and Mark Konishi and Paul Patterson for helpful criticism of the manuscript. This research was supported by grants to S. B. from the National Science Foundation (BCS-8908154),the National Institutes of Health (GM 40499 and EYO9278), and the McKnight Endowment Fund for Neuroscience. R. L. B. was supported by a National Research Service Award (5F32NS0881-02) from the National Institute of Neurological Disorders and Stroke and a fellowship from the French Foundation for Alzheimer's Research. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC Section 1734 solely to indicate this fact. Received July 14, 1992; revised December 22, 1992.

References Anderson, D. J. (1989). New roles for PDGF and CNTF in controlling the timing of glial cell differentiation in the optic nerve. Trends Neurosci. 12, 83-85. Ashburner, M. (1989). Drosophila: A Laboratory Manual (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory), pp. 179-189. Barres~B. A. (1991). New roles for glia. ]. Neurosci. 11, 3685-3694. Benzer, S. (1971). From gene to behavior. J. Am. Med. Assoc. 218, 1015-1022. Bloomquist, B. T., Shortridge, R. D., Schneuwly, S., Perdew, M., Montell, C., Steller, H., Rubin, G., and Pak, W. L. (1988). Isolation of a putative phospholipase C gene of Drosophila, norpA, and its role in phototransduction. Cell 54, 723-733. Boschek, C. B. (1971). On the fine structure of the peripheral retina and lamina ganglionaris of the fly Musca domestica. Z. Zellforsch. Mikrosh. Anat. 118, 369-409. Carlson, S. D., and St. Marie, R. L. (1990). Structure and function of insect glia. Annu. Rev. Entomol. 35, 597-621. Carthew, R. W., and Rubin, G. M. (1990). seven in absentia, a gene required for specification of R7 cell fate in Drosophila eye. Cell 63, 561-577. Coombe, P. E., and Heisenberg, M. (1986). The structural brain mutant Vacuolar medulla of Drosophila melanogaster with specific behavioural defects and cell degeneration in the adult. J. Neurogenet. 3, 135-158. Driscoll, M., and Chalfie, M. (1992). Developmental and abnormal cell death in C. elegans. Trends Neurosci. 15, 15-19. Dryja, T. P., McGee, T. L., Reichel, E., Hahn, L. B., Cowley, G. S., Yandell, D. W., Sandberg, M. A., and Berson, E. L. (1990). A point mutation of the rhodopsin gene in one form of retinitis pigmentosa. Nature 343, 364-366. Duvall, E., and Wyllie, A. M. (1986). Death and the cell. Immunol. Today 7, 115-119. Fischbach, K. F., and Dittrich, A. P. M. (1989). The optic lobe of Drosophila melanogaster. I. A Golgi analysis of wild-type structure. Cell Tissue Res. 258, 441-475. Fischbach, K. F., and Technau, G. (1984). Cell degeneration in the developing optic lobes of the sine oculis and small-optic-lobes mutants of Drosophila melanogaster. Dev. Biol. 104, 219-239. Flanagan, J. R. (1977). A method for fate mapping the loci of lethal and behavioral mutants in Drosophila melanogaster. Genetics 85, 587-687. Freeman, M., Kl~mbt, C., Goodman, C. S., and Rubin, G. M. (1992). The argos gene encodes a diffusible factor that regulates cell fate decisions in the Drosophila eye. Cell 69, 963-975.

Brain Degeneration Mutant 849

Gasser, U. E., and Hatten, M. E. (1990). Neuron-glial interactions of rat hippocampal cells in vitro: glial guided neuronal migration and neuronal regulation of glial differentiation. J. Neurosci. 10, 1276-1285. Grenningloh, G., Bieber, A. J., Rehm, E. J., Snow, P. M., Traquina, Z. R., Hortsch, M., Patel, N. M., and Goodman, C. S. (1990). Molecular genetics of neuronal recognition in Drosophila: evolution and function of immunoglobulin superfamily cell adhesion molecules. Cold Spring Harbor Symp. Quant. Biol. 55, 327-340. Griffiths, G. W. (1979). Transport of glial cell acid phosphatase by endoplasmic reticulum into damaged axons. J. Cell Sci. 36, 361-389. Grigliatti, T. (1986). Mutagenesis. In Drosophila: A Practical Approach (Oxford: DCIRL Press), pp. 39-58. Heisenberg, M. (1980). Mutants of brain structure and function: What is the significance of the mushroom bodies for behavior? In Development and Neurobiology of Drosophila, O. Siddiqi, P. Babu, L. M. Hall, and J. C. Hall, eds. (NewYork: Plenum Publishing Corp.), pp. 373-390. Henkemeyer, M. J., Gertler, F. B., Goodman, W., and Hoffman, F. M. (1987). The Drosophila Abelson proto-oncogene homolog: identification of mutant alleles that have pleiotropic effects late in development. Cell 51, 821-828. Hofbauer, A., and Campos-Ortega, J. A. (1990). Proliferation pattern and early differentiation of the optic lobes in Drosophila melanogaster. Roux's Arch. Dev. Biol. 198, 264-274. Homyk, T., Sinclair, D. A. R., Wong, D. T. L., and Grigliatti, T. A. (1986). Recovery and characterization of temperature-sensitive mutations affecting adult viability in Drosophila melanogaster. Genetics 113, 367-389. Hotta, Y., and Benzer, S. (1970). Genetic dissection of the Drosophila nervous system by means of mosaics. Proc. Natl. Acad. Sci. USA 67, 1156-1163. Hotta, Y., and Benzer, S. (1972). Mapping of behaviour in Drosophila mosaics. Nature 240, 527-535. Hoyle, G. (1986). Glial cells of an insect ganglion. J. Comp. Neurol. 246, 85-103. Inoue, H., Yoshioka, T., and Hotta, Y. (1989). Diacylglycerol kinase defect in a Drosophila retinal degeneration mutant rdgA. J. Biol. Chem. 264, 5996-6000. Jacobs, J. R., Hiromi, Y., Patel, N. H., and Goodman, C.S. (1989). Lineage, migration, and morphogenesis of longitudinal glia in the Drosophila CNS as revealed by a molecular lineage marker. Neuron 2, 1625-1631. Kankel, D. R., and Hall, J. C. (1976). Fate mapping of nervous system and other internal tissues in genetic mosaics of Drosophila melanogaster. Dev. Biol. 48, 1-24. Kankel, D. R., Ferrus, A., Garen, S. H., Harte, P. J., and Lewis, P. E. (1980). The structure and development of the nervous system. In The Genetics and Biology of Drosophila, Volume 2 (New York: Academic Press), pp. 295-363. Knapp, P. E., Dutta, S., and Skoff, R. P. (1990). Differences in levels of neuroglial cell death in Jimpy male mice and carrier females. Dev. Neurosci. 12, 145-152. Kuffler, S. W. (1967). Neuroglial cells: physiological properties and a potassium mediated effect of neuronal activity on the glial membrane potential. Proc. R. Soc. Lond. (B) 168, 1-21. Lane, N. J. (1985). Structure of components of the nervous system. In Comparative Insect Physiology, Biochemistry, and Pharmacology, Volume 5, G. A. Kerkut, L. I. Gilbert, eds. (Oxford: Pergamon Press), pp. 1-47. Lindsley, D. L., and Zimm, G. G. (1992).The Genome of Drosophila melanogaster (New York: Academic Press). McKusick, V. A. (1990). Mendelian Inheritance in Man (Baltimore: The Johns Hopkins University Press). Meyer, M., Reddy, G. R., and Edwards, J. S. (1987). Immunological probes reveal spatial and developmental diversity in insect neuroglia. J. Neurosci. 7, 512-521.

Miller, A. (1965).The internal anatomy and histology of the imago of Drosophila melanogaster. In Biology of Drosophila, M. Demerec, ed. (New York: Hafner Publishing Co.), pp. 420-531. Montell, C., and Rubin, G. M. (1989). Molecular characterization of the Drosophila trp locus: a putative integral membrane protein required for phototransduction. Neuron 2, 1313-1323. Nambu, J. R., Franks, R. G., Hu, S., and Crews, S. T. (1990). The single-minded gene of Drosophila is required for the expression of genes important for the development of CNS midline cells. Cell 63, 63-75. Nordlander, R. H., and Edwards, J. S. (1969). Post-embryonic brain developm ent in the monarch butterfly, Danaus plexippus plexippus (L). Roux's Arch. Dev. Biol. 162, 197-217. O'Tousa, J. E., Leonard, D. S., and Pak, W. L. (1989). Morphological defects in orajK84 photoreceptors caused by mutations in R1-6 opsin gene of Drosophila. J. Neurogenet. 6, 41-52. Padgett, R. W., St. Johnston, R. D., and Gelbart, W. R. (1987). A transcript from a Drosophila pattern gene predicts a protein homologous to the transforming growth factor-13family. Nature 325, 81-84. Papp, M. I., Kahn, J. E., and Lantos, P. L. (1989). Glial cytoplasmic inclusions in the CNS of patients with multiple system atrophy (striatonigral degeneration, olivopontocerebellar atrophy and Shy-Drager syndrome). J. Neurol. Sci. 94, 79-100. Pentreath, V. W. (1989). Invertebrate glial cells. Comp. Biochem. Physiol. 93,4, 77-83. Ramon y Cajal, s. R., and Sanchez, D. (1915). Contribuci6n al conocimiento de los centros nerviosos de los insectos. Trab. Lab. Invest. Biol. Univ. Madr. 13, 1-168. Renfranz, P. J., and Benzer, S. (1989). Monoclonal antibody probes discriminate early and late mutant defects in development of the Drosophila retina. Dev. Biol. 136, 411-429. Reynolds, E. S. (1963). The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17, 208-212. Rothberg, J. M., Jacobs, J. R., Goodman, C. S., and ArtavanisTsakonas, S. (1990). slit: an extracellular protein necessary for development of midline gila and commissural axon pathways contains both EGF and LRR domains. Genes Dev. 4, 2169-2187. Siman, R., Card, J. P., Nelson, R. B., and Davis, L. G. (1989). Expression of I~-amyloid precursor protein in reactive astrocytes following neuronal damage. Neuron 3, 275-285. Steele, F., and O'Tousa, J. E. (1990). Rhodopsin activation causes retinal degeneration in Drosophila rdgC mutant. Neuron 4, 883890. Stewart, B. R., and Merriam, J. R. (1975). Regulation of gene activity by dosage compensation at the chromosome level in Drosophila. Genetics 79, 635-647. Stocker, R. F., Edwards, J. S., and Truman, J. W. (1978). Fine structure of degenerating abdominal motor neurons after eclosion in the sphingid moth, Manduca sexta. Cell Tissue Res. 191, 317331. Strausfeld, N. J. (1976). Atlas of an Insect Brain (Berlin: SpringerVerlag). Strausfeld, N. J. (1980). The Golgi method: its application to the insect nervous system and the phenomenon of stochastic impregnation. In Neu roanatomical Techniques, N. J. St rausfeld and T. A. Miller, eds. (Berlin: Springer-Verlag), pp. 132-190. Tel, H., Nihonmatsu, I., Yokokura, T., Ueda, R., Sano, Y., Okuda, T., Sato, K., Hirata, K., Fujita, S., and Yamamoto, D. (1992). pokkuri, a Drosophila gene encoding an Ets domain protein, prevents overproduction of the R7 photoreceptor. Proc. Natl. Acad. Sci. USA, in press. Tolbert, L. P., and Oland, L. A. (1990). Glial cells form boundaries for developing insect olfactory glomeruli. Exp. Neurol. 109, 1928. Trujillo-Cen6z, O. (1965). Some aspects of the structural organization of the intermediate retina of dipterans. J. Ultrastruct. Res. 13, 1-33.

Neuron 850

Trujillo~Cen6z, O. (1985). The eye: development, structure and neural connections. In Comparative Insect Physiology, Biochemistry, and Pharmacology, Volume 6, G. A. Kerkut and L. I. Gilbert, eds. (Oxford: Pergamon Press), pp. 171-223. Tsacopoulus, M., EvC~lUoz-Mercier , V., Perrottet, P., and Buchner, E. (1988). Honeybee retinal glial cells transform glucose and supply the neurons with metabolic substrate. Proc. Natl. Acad. Sci. USA 85, 8727-8731. Vihtelic, T. S., Hyde, D. R., and O'lousa, J. E. (1991). Isolation and characterization of the Drosophila retinal degeneration B (rdgB) gene. Genetics 127, 761-768. Zuker, C. S., Cowman, A. F., and Rubin, G. M. (1985). Isolation and structure of a rhodopsin gene from D. melanogaster. Cell

40, 851-858.