Copyright 0 1996 by the Genetics Society of America
Drosophila syntaxin Is Required for Cell Viability and May Function in Membrane Formation and Stabilization Karen L. Schulze*” and Hugo J. Bellen*,+ *Division of Neuroscience and +Department of Molecular and H u m a n Genetics, +Department of Cell Biology, Howard Hughes Medical Institute, Baylor College of Medicine, Houston, Texas 77030 Manuscript received June 25, 1996 Accepted for publication September 17, 1996
ABSTRACT The role of the Drosophila homologue of syntaxin-lA (syx) in neurotransmission has been extensively studied. However, developmental Northern analyses and in situ hybridization experiments show that SYX mRNA is expressed during all stages and in many tissues. We have isolated new mutations in syx that reveal roles for syx outside the nervous system. In the ovary, SYX is present in the germarium, but it is predominantly localized to nurse cell membranes. Mitotic recombination experiments in the germline show SYX is essential for oogenesis and may participate in membrane biogenesis in the nursecells. In the early embryo, a large contribution of maternally deposited RNAis present, and the protein is localized at cell membranes duringcellularization. After the maternal contribution is depleted, zygotically produced SYX assists secretion events occurring late in embryogenesis, such as cuticle deposition and neurotransmitter release. However, SYX is also required in larval imaginal discs, as certain hypomorphic mutant combinations exhibit rough eyes and wing notch defects indicative of cell death. Furthermore, recombinant clones that lack syx cause cell lethality in the developing eye. We propose that, similar to its roles in cuticle secretion and neurotransmitter release, SYX may mediate membrane assembly events throughout Drosophila development.
T
HE study of vesicle-mediated transport in a number of systems has greatly enhanced our understanding of the process by which membranes and their protein constituents are delivered to the proper targetsite during secretion and membrane addition. The observation that the molecular mechanisms that underlie vesicle trafficking in two distinct systems, constitutive secretion in yeast and regulated neurotransmitter release in higher eukaryotes, are conserved has led to the hypothesis that an ubiquitous machinery exists to execute the fusion event (reviewed in BENNETTand SCHELLER 1993; FERRO-NOVICK andJmN 1994). Extensive studies of the molecules that are proposedto mediate vesicle docking and fusion events in both systemshave generated a number of theories regarding how these events occur. The model that has enjoyed the most popularity is the “SNARE hypothesis.” The basic premise relies upon the interaction of putative membrane-anchored receptors, termed v- (vesicle) and t- (target) SNARES,dependent on the membrane within which they reside. The SNAREs (SNAPreceptors) are so named for theirability to attract the soluble NSF attachment proteins(SNAPS) and theATPase NSF (Nethylmaleimide-sensitive fusion protein) to the appropriate position to effect vesicle Corresponding uuthor: Hugo J. Bellen, Howard Hughes Medical Institute, Baylor College of Medicine,RoomT630, One Baylor Plaza, Houston, TX 77030. E-mail:
[email protected] Present address: Howard Hughes Medical Institute, University of California, 5-748 MRL, P.O. Box951662, Los Angeles, CA 900951662.
’
Genetics 144: 1713-1724 (December, 1996)
fusion with its target membrane (SOLLNERet al. 1993a,b). In brain, theSNAREs have been identified as synaptobrevin on the synaptic vesicle and syntaxin and SNAP-25 withinthe presynaptic membrane. Theseproteins have homologues in yeast that have independently been identified as essential genes that lie alongthe secretory pathway. It has therefore been speculated that the association of tissue- or cell-specific vesicleand target receptors with each other and/or with the NSFSNAP complex may mediate the docking and fusion event (SOLLNER et al. 1993b; PEVSNER et al. 1994; SCHIAVO et al. 1995). The phenotypes of yeast SNARE mutants (secretion blockade and post-Golgivesicle accumulation)confirms the absolute necessity for these proteins in exocytosis(NOVICK et al. 1980).Furthermore,thebrain SNARE homologues have also been shown to be essential to neurotransmitter release, as they are the targets of proteolytic cleavage by the clostridial and botulinum neurotoxins that block neurotransmission (SCHIAVO et al. 1992;HAYASHIet al. 1994; reviewed in JAHN and NIEMANN1994). Though these data certainly do not contradict the premises of the theory, recent evidence has challenged the fundamental tenets of the SNARE hypothesis. Mutational analyses of two proteins, the putative v-SNARE n-synaptobrevin andthepotential tSNARE syntaxin, have been performed in Drosophila (SWEENEY et al. 1994; BROADIE et al. 1995; SCHULZE et al. 1995). These experiments have established that synaptobrevin and syntaxin bothfunction downstream
1714
K. L. Schulze and H. J. Bellen
(TEPASS et al. 1990). Some syx alleles were rebalanced over TM6B and placed in a y w background using either the balancer strain y w/y w; D3/TM6B, T6 P(w+, AbdA-lacZJ or y w/y w; L/CyO; P{ry’, 1acZJslp;@/TMGB, TbP[w’, A b d - l a c Z ) . The P-element insertion strainP(ly+tsyx” t y / TM3, SI, ry (abbreviated syx”) was obtained from the Drosophila Stock Center in Bloomington, Indiana (P[syx], SCHULZE et uL. 1995). Excision of the syx‘insertion was achieved by mating individual P[ry+]syx~/TTM?males with virgin females of the genotype y w/y w; Ki p” A2-3/Ki p‘ A2-3. Individual y w; P(ly+)syx”y/ Ki A2-3 males (or females) were backcrossed to females (males) of the original insertion strain to balance the excision chromosome. The progeny of these crosses were scored for individuals with rosy eye color, indicative of an excision event, from each of which Asyx‘ ry/ TM?, Sb ly balanced stocks were established. Complementation tests withthe original insertion strain were subsequently performed to define the nature of the excision events as precise or imprecise. To isolate point mutations, isogenized ry emales were fed 25 mM EMS suspended in 1% sucrose using a standard protocol (LEWISand BACHER1968). The mutagenized chromosomes (10,500) were balanced over TMGB, Tband assessed for failure to complement the original Pelement insertion syx” at 28”. Induction of mosaic animals and germline clones: Genetic mosaics were generated using the FLP/FRT recombinase system (CHOU and PERRIMON 1992) essentially as outlined by XU and RUBIN (1993). The alleles ~yx’’~~’, syx‘ (Df(?R)A229 and Df(?R)A6, SCHULZE et al. 1995) and syx’ (each in a y 711 background) were recombinedonto a chromosome containing Ply+ hs-neo FRT}82B using selection for resistance to G418 (Geneticin, GIBCO) and verification of failure to complement theoriginal J-yxallele. Malescontaining theheatinducible FLP recomhinase (y w P/ly+ hsFLPI/Y) andthe chromosomes Plry* hs-neo FRT/82B P{w’ ly’}9OE (permitting detection of mosaic clones in the eye) or Plry’ hs-neo FRTjr82B P{w+a v d ” / ~P{U+ ~ OVO”‘/?RZ (for assaying germline clone development) were constructed and crossed to y 7u; P(y+ hsneo FRT)82B syx females. First instar larvae (24-48 hr AEL) were incubated in a 38” water bath (in the vials in which they were laid) for 1 hr to induce mitotic recombination. Adult females of theappropriate genotype were examined for clones ( w + ) or assayed for egg-laying ability after several days of mating to y w; syxnZz9/TM6B,Tb males. In some cases, the ovaries of females that failed to lay eggs were dissected (see below). P(FRT} and P(rnarker] strains were obtained from the Bloomington, Indiana Stock Center. Construction of genomicrescuetransgenes: An 11.O-kb XbaI fragment from genomic phage A10 and a 6.0-kb NotI fragment from genomic phage A6 (see Figure 3) were subcloned into the vector pCaSpeR3, which contains the 5’ and 3’ P-element integration sequences flanking the white gene (PIRROTTA 1988). The constructs were injected into y w; Ki ppA2-3/+ embryos (BELLENet al. 1992), andsurviving adults were backcrossed to y w. Progeny bearing w+ eyes (indicating stable integration of the construct at a chromosomal locus) were selected and served as founders of stocks. Seven second chromosome insertion strains were generated that contained the A10 X6aI construct, and six second chromosome insertion strains were created that bore the A6 NotI construct. To determineif either of the two P[w+lsyx’ constructs could rescue the lethality of syx‘ and ~ y x ”y ~w; ~P[w’lsyx+/CyO; ~, D/ TM6B, Tb males were crossed to y w; syx-/TM63, TI,females. Of the progeny of this cross, brothers and sisters of the genoMATERIALS AND METHODS type y w; P(w+)syx+/+;syx-/TMGB, Tb were mated. Recovery of non-Tubby progeny was indicative of rescue. Drosophila stocks and mutagenesis: All Drosophila strains In situ hybridization: To determine the distribution of SYX were maintained on standard cornmeal-molasses medium at mRNA in wild-type embryos, antisense RNA probes were syn25” unless otherwise noted. Canton-Swasused as the wild-type strain. A deficiency that uncovers all g x alleles (Df?R)ml~’~~~, thesized from the full-length 1.4kb cDNA subcloned into Bluescript I1 SK+. RNA digoxigenin-labeled probes were prebreakpoints 95D7- 11; 95F15) was provided by Dr. EIJZABETH
of the vesicle docking event, in contrast to what was originally proposedinthe SNARE hypothesis. More importantly, genetic manipulations such as these that removed N-SYB or SYX function revealed that although synaptobrevin may assist in Cay’-regulated neurotransmitter release, its function is not essential to the fusion event per se as spontaneous fusions still occur (SWEENEY et al. 1994; BROADIE et al. 1995). On the other hand, the role of syntaxin was shown to be absolutely essential to vesicle fusion in neurotransmitterrelease, as allforms of exocytosis, both spontaneous and evoked, are absent in syx null mutants (BROADIEet al. 1995; SCHULZE et al. 1995). The precise function of syntaxin, whether in a prefusion or “priming” step, or as a catalystof the fusion event itself, has yet to be elucidated. Clues to syntaxin’s function have been revealed by rigorous analysis of its protein structure and its interactions with other proteins. Syntaxin has been shown to bind in vitro to all the components that may form the SNARE complex (synaptobrevin, SNAP-25, NSF and aSNAP), and amino acid domains within the protein that are crucial for these interactions have been delineated (CMOS et al. 1994; HAYASHI et al. 1994; PEVSNER et al. 1994; HANSON et al. 1995; K E E et al. 1995). Syntaxin was originally isolated due to its interaction with the synaptic vesicle membrane protein synaptotagmin, and this interaction has been shown to occur in a Ca‘+-dependentmanner (BENNEMet al. 1992; CHAPMAN et al. 1995). Finaily, syntaxin also interacts with presynaptic N-type calcium channels (SHENGet aE. 1994) in a manner that may effect their function (BEZPROZVANNY et al. 1995). The association of syntaxin with synaptotagmin and Ca2+channels in the neuron strongly supports the placement of syntaxin’s function late in the cascade of events that culminatein vesicle fusion upon Ca2+influx. We initially reported that the expression of the Drosophila syntaxin-lA homologue is not restricted to the nervous system as is thought to be the case for its vertebrate counterpart.syxfunctions in cuticle secretion and possibly in the garland cells where it may assist in the clearance of waste from the hemolymph via endo- and exocytosis (SCHULZE et al. 1995). Togain further insight into SYNTAXIN’S( S Y X ) potential role in vesicle fusion, we have examined its function in a variety of secretory processes occurringthroughout Drosophila develop ment. Hereinwe present preliminary evidence that SYX may be required for many exocytotic events during fusion or prefusion steps that may occur in a similar manner as has been suggested for neurotransmission. Cumultatively, these data point to a critical and essential role for SYX in membrane fusion and biogenesis.
KNUST
6’
Role of q x in Membrane Biogenesis
1715
for each mutant. The Genetics Computer Group Sequence pared as described in theGenius RNA nonradioactive labeling Analysis Package Version 7.3.1-UNIX was utilized for all subsedetection kit (Boehringer Mannheim) . In situ hybridization to whole mount embryos was performed as previously dequent nucleotide and amino acid sequence analyses. The molecular defects of excision alleles were determined scribed by INGHAMet al. (1991) and TAUTZand PFEIFLE by performing Southern analyses of genomic DNA prepared (1989). from balanced strains carried out essentially as described in Immunobistochemis~Appropriately staged wild-type SAMBROOK et al. (1989). PCR analysis was performed on crude and syx mutant embryos were collected and processed for genomic DNA of single flies using the 31-bp inverted P-eleimmunocytochemicalstaining according to standard techment repeat (5'-CGACGGGACCACClTATGTTATTTCATniques outlined in SALZBERG et al. (1994). MAb 8C3, which CATG-3') as well as two primers designed from syx sequence recognizes Drosophila Syntaxin-1A and was given by Dr. SEY(5'GTTATTTGTGACTCACACTGGTACC-3' and 5'-GTTMOUR BENZER, was used at a final dilution of 1:10 (FUJITAet CACACTCGAAAATTGTITGCCC-3') to determine if the P al. 1982). MAb 1D4, which recognizes the cytoplasmic domain element had been completely removed or internally deleted of Drosophila Fasciclin I1 (VANVACTORet al. 1993), was o b (see SCHULZE et al. 1995). The conditions for the PCR were tained from Dr. COREYS. GOODMAN and was used at a final as follows: 1 cycle at 95" (5 min); 35 cycles at 94" (1.5 min), dilution of 1:50 to determine if motorneuron growth cone 51-53" (2 min) and 72" (3 min); 1 cycle at 72" (3 min). guidance and fasciculation occurred normally. MAb 22C10, a neuronal marker that recognizes a membrane antigen on all cells of the Drosophila peripheral nervous system (FUJITA RESULTS et al. 1982; GOODMAN et al. 1984) was utilized at a 1:200 dilution to analyze overall neuronal morphology. Nuclear staining y x expression is widespread throughout Drosophila of embryos was achieved by treating fixed embryos with 125 development: We initially reported the expression and units ribonuclease A for 30 min at 37". Propidium iodide (50 distribution of SYX transcripts and SYX protein during pg/mL) was applied concurrently with the primary antibody. the latter half of embryogenesis (SCHULZE et al. 1995). Ovaries were dissected and fixed as described in XUE and COOLEY (1993). Fixed ovaries were processed for immunohisThe extent of expression of SEX during Drosophila detochemistry using theprocedurefor embryos outlinedin velopment was further investigated by Northern blotSALZBERG et al. (1994). To visualize SYX protein distribution, ting of RNA representing all stages of embryogenesis, MAb 8C3 was utilized at a dilution of 1:10. Biotinylated goat larval life and adulthood. We tested three probes: the anti-mouse secondary antibody (Vector) was used at a 1:200 3' end of the cDNA containingcoding region that dilution. The Vecta-Stain ABGHRP kit (Vector) was used to would be most conserved (C-terminal portion of the augment the DAB-peroxidase signal. Fluorescein isothiocyanate-conjugated anti-mouse IgG was utilized at a 1:200 diluprotein), a probe containing 5'untranslated sequence tion. and sequence coding for the first 44 amino acids (not Paraffin sections of adult Canton-S heads were prepared for well conserved among syntaxins), andaprobe conmass histology using a modification of the protocol ofJAGER taining only 3' untranslated sequence. All three probes and FISCHBACH(1987).The primary antibody (anti-rat recognized the same complex pattern composed of up Syntaxin polyclonal 1378, HATA et al. 1993) was diluted 1:200. Molecular techniques:Total RNA for Northernanalysis was to six transcripts, ranging in size from 3.5 to 12 kb (see isolated by LiCl precipitation of phenol-chloroform extracted Figure 1). SYX messages are present during every stage Drosophila homogenates, andpoly-A+ RNA was purified with of Drosophila development. Although the pattern of oligodT cellulose ( SAMBROOK et al. 1989). Approximately 5 transcripts isvery complex considering that the small pg poly-A+ RNA was loaded per lane. The probe was gener(1.4 kb) cDNA we and others (CEREZO et al. 1995) have ated by PCR of the 3' untranslated region of the syx cDNA, including 430 bp between the stop codon and the polyadenylrecovered is contained within a single exon, it is likely ation signal. A 3.0-kb EcoRI genomic fragment served as the that alternative splicing, various promoters, and differtemplate for thereaction. The primers usedwere 5'-ClTACAent polyadenyltion signalsare used to generate this hetCGCCACATTCACTCC-3' and 5"CTTTGTTTGCAATTAerogeneity. We feel confident that each of these tranCGTGTGJ' with the following conditions: 1 cycle at 95" (5 scripts represent SYXmessages and nottranscripts from min); 35 cycles at 94" (1.5 rnin), 51" (2 min) and72" (3 min); 1 cycle at 72" (3 min). Other probes tested were the 3' half other related and possibly highlyhomologous Drosophof the cDNA from the internalEcoRI site to the end (encodes ila syntaxinsthat may exist, since three different probes last 174 amino acids) and a 330-bp HindIII/SacII fragment generated identical patterns. Further, these transcripts from the 5' half of the cDNA (includes codons of the first 44 are likely to encode a single protein as developmental amino acids). Westerns displayonly a single band(SCHULZEet al. To determine themolecular defects of EMSinduced point mutants, crude genomic DNA was prepared from individual 1995; data not shown). In addition, low stringency hy~yx'/syx""~or syx'/syx' embryos or single male s y ~ ' ~ ~ ~ / sbridization y ~ ~ ~ ~of~ genomic Southern blots failed to reveal flies to serve as a template for PCR amplification of the open any other bands than those corresponding to SYX (data readingframe (ORF) using complementaryprimers with not shown). overhangs containing BamHI and Hind111 restriction sites to In young embryos [0-3 hr after egg lay (AEL)], the facilitate subcloning. The conditions for the PCR were as fol3.5-, 4.2- and 8.0-kb transcripts are abundant. At 3-9 lows: 1 cycle at 95" (5 min); 35 cycles at 94" (1.5 min), 57" (2 min), 72" (3 min);1 cycle at 72" (3 min). The productof the hr AEL, the 3.5-kb transcript slowly disappears while reactions were subcloned intoBluescript I1 SK+ (Stratagene), the 4.2-, 8.0-and 9.0-kb messagespredominate until the and the universal and reverse primers as well as a panel of end of embryogenesis. Between 9-15 AEL,two addiinternal primers were utilized to prime sequencing reactions tional transcripts (7.0 and 12 kb) appear. During first by the dideoxy chain-termination method using the autoinstar larval life, five messages are present, whereas secmated fluorescent procedure (Applied Biosystems). Two clones independently isolated in this manner were sequenced ond and third instars utilize three and two messages,
K. I,. Schulze and H. J. Bcllcn
1716
kb
412 + 9.0 * 8.0 * 7.0
9.4.
7.4. 4.4
44
2.4
L
4.2 3.5
RP49
FI(:L‘RI.: l.--Dc\c~lop~~~c~lt“1 csprcwiion o f .syx. Six principal s ~ transcripts x (ranging in size from 3.5 to 12 kb) are present rlrlring Drosophila development. The Northern blot was hy-
bridized w i t h a specific . y probe amplified by PCR from the 3‘ untl-anslatetl region. All the lanes were equally loaded based o n hyhridization of RP49 probe, with the exception of secondinstars, pupae and adults that are slightly underloaded.
respectively. During pupariation, all five transcripts reappear, and the adult expression pattern resembles that of the pupa. MTeobserved that three of the adult transcripts (3..5,4.2 and 8.0 kb) arealso present in 0- to 9 h r old embryos, suggesting some or all of these messages represent the maternal contribution to the embryo. We continuedthedevelopmental analysis by performing in situ hybridization and immunohistochemisty to more closely examine syx’s expression in young embryos.Digoxigenin-labeledantisense RNA probes were utilized to visualize the distribrltion of SYXmRNA in embryos of a11 stages. In agreement with the Northern analysis, which indicated three prevalent transcripts in embryos aged 0-3 h r AEL (Figure l ) , very young embryos contain high levels of syntaxin RNA and protein (Figure 2). Precellalarized embryos exhibit ubiquitously distrik uted syx mRNA, although it accumulates most densely in the germ plasm at the extreme posterior tip of the embryo where the pole cells will form (Figure 2B). As the future polecell membranes pinchoff from the posterior tip, SJ’X RNAis located within each bud and remainsconcentratedattheposterior of theembryo (data not shown). Upon the completion of somatic cellularization, syx RNA is uniformly distributed within all cells of the embryo and seems concentrated near the cell membrane (data not shown). As gastrulation proceeds, s y RNA ~ is abundant in all cells. At stage 9, the message is present in most ectodermal cells, the anterior and posterior midgut invaginations, and the central nervous system (CNS) precursor cells (Figure 2F). Expression of syxduring laterstages of embryogenesis was previously characterized and includes the garlandcells, midgut, nervous system and ectoderm (SCHULZE rt nl. 199.5).
To determine the distribution of SYX protein during oogenesis and embryogenesis, we immunocytochemically stained ovaries and embryos with either a polyclonal antibody raised against rat syntaxin-lA(HATArt 01. 1993) o r monoclonal antibody 8C3 (a gift from KOSRXI) ZINS Mr\IE:R and SI?Y%IOL!R BENZER). The former antibodyspecifically recognizes a protein of the appropriatemolecular weight in extracts prepared from Drosophila heads and identifies theproteinencoded by the .yx gene (SCHUI.7.E rt nl. 1995), whereas the latter was prepared against the native Drosophila SYX protein ( F ~ ) ~ IrtT A nl. 1982). In wild-type ovaries stained withMAb 8C3, we detected SYX protein in regions 2 and 3 of the germarium, outlining the membranes of germline cyst cells. SYX continues to be abundantly expressed in the nurse cell membranes of egg chambers duringstages 1-8, and SYX levels fade during stages 8 and 9 (Figure 2A). In young embryos during mitotic cycles 9-13, the plasma membrane partially envelops each dividing nncleus, forming “cytoplasmic buds”. Beginning inmitotic cycle 9, “pole buds” form at the posterior tip of the embryo, marking the formation and segregationof the germ cells (FOE P t nl. 1993). As is the case for syx mRNA, SYX is present ubiquitously in precellularized embryos but is concentrated at the posterior tip, and accumulates beneath these pole buds as they undergo cytokinesis (Figure 2, C and D). Similarly during cycle 11 “somatic buds” form around dividing somatic nuclei but undergo several rounds of formation and collapse as the nuclei beneath the budsdivide. At metaphase of each mitotic division, the membrane of the bud invaginates markedly, and “pseudocleavagefurrows”form betweenthebuds (FOE rt nl. 1993; SCHF..TER and M‘IESCHAUS 1993). SYX expression appearsto follow the outline of the budsas they partiallyenwrap thedividing nuclei during mitotic cycles 10-14 (Figure 2C).Finally, during cycle 14, the somatic nuclei become completely surrounded by the elongating membranes of the cytoplasmic buds, and cellularization is completed. SYX Itr calization outlines these membranes as they form (Figure2D). SYX continues to be expressed in the membranes of each cell of the fully cellularized e m h r y (Figure 2E). Syntaxin proteinexpressiondoesnotbecome restricted to any tissue o r cell type until stage 9 of embryogenesis. During germ band extension, the protein is present in the mesoderm, ectoderm and invaginating neurectoderm but is excluded from the amnioserosa (data not shown). During later stages of embryogenesis, the protein is most prominent in the CNS and the garland cells and is weakly expressed in the midgut and ectoderm (SCHULLE rt 01. 1995). SYX protein is also present in the adult brain and the synaptic substationsof the visual system (Figure 5A). Staining of paraffinsections through adult wild-type heads with an anti-rat syntaxin polyclonal antibody reveals SYX is concentrated in the neuropil regions of the brain andis somewhat enriched at synaptic regions,
Role of syx in Membrane Biogenesis
1717
FIGURE 2.-Distribution of SYXmRNA and SYX protein in ovarioles and young embryos. Wild-type ovariole and embryos stained for SYX with MAb 8C3 or hybridized with an antisense SYXRNA probe. Embryos are oriented anterior to the left, dorsal up. Staging is according to CAMPOS-~RTEGA and HARTENSTEIN (1985). (A) Syntaxin is present in the nurse cell membranes of the ovary. The protein is abundant in the nurse cell membranes of the germarium (g) andthe early stages of oocyte development but dissipates beginning at stage 8 (S8). (B)SYXRNA ispresent ubiquitously in embryos before cellularization but is concentrated at the posterior pole. (C) SYX distribution in an embryo that has not yet completed mitotic cycle 13 is observed in large, roughly hexagonal-shaped outlines that correspond to the pseudocleavage furrows of the cytoplasmic buds. (D) An embryo that nears the completion of somatic cellularization (cycle 14) demonstrates SYX staining along the membranes of each cell (green). SYX is also concentrated beneath the pole cells. Nuclei are labeled with propidium iodide (red). (E) A completely cellularized embryo is beginning to undergo gastrulation (stage 6), as the cephalic and ventral furrows are forming. SYX is present in all the membranes of all cells at this stage. (F) A stage 9 embryo exhibits SYXRNA localization in the anterior (am) and posterior (pm) midgut invaginations, the ventral neuroblasts (n) and the ectoderm.
such as in the lamina and medulla of the optic lobes where photoreceptor axons form synaptic contacts with second order neurons. However, SYX distribution in the brain differs from that of synaptic vesicle-specific proteins suchas synaptotagmin (LITTLETON et al. 1993), as SYX is also present in axonsand cell bodies whereas synaptotagmin is restricted to synaptic terminals. Generation and molecular characterization of muta-
tions in syx: To gain a better insightinto syx's function, and to identify possible new phenotypes associated with partial loss of function alleles of syx, we performed two mutagenesis screens. In the first, the P element in ?xp was impreciselyexcised (SCHULZE et al. 1995). This P element created an hypomorphic syx mutation that compromisedneurotransmissionat the embryonicneuromuscularjunction. Mature ?xp embryos failedto exhibit
1718
K. L. Schulze and H. J. Bellen
muscle contraction waves and emerge from the egg case. within the Pelement (syx', syx", syx ' I y and syxZy4).HowIn addition, two of the embryonic lethal mutations recovever, a third group, represented by alleles with
