A Genetic Screen for Synaptic Transmission Mutants Mapping to the

Copyright 2003 by the Genetics Society of America

A Genetic Screen for Synaptic Transmission Mutants Mapping to the Right Armof Chromosome 3 in Drosophila

Michael C. Babcock,*,1,2 R. Steven Stowers,†,1,3 Jennifer Leither,*,2 Corey S. Goodman†,4

and Leo J. Pallanck*,5

*Department of Genome Sciences, University of Washington, Seattle, Washington 98195-7730 and †Department of Molecular and Cell Biology,University of California, Berkeley, California 94720

Manuscript received March 5, 2003Accepted for publication April 25, 2003

ABSTRACTNeuronal function depends upon the proper formation of synaptic connections and rapid communica-

tion at these sites, primarily through the regulated exocytosis of chemical neurotransmitters. Recentbiochemical and genomic studies have identified a large number of candidate molecules that may functionin these processes. To complement these studies, we are pursuing a genetic approach to identify genesaffecting synaptic transmission in the Drosophila visual system. Our screening approach involves a recentlydescribed genetic method allowing efficient production of mosaic flies whose eyes are entirely homozygousfor a mutagenized chromosome arm. From a screen of 42,500 mutagenized flies, 32 mutations on chromo-some 3R that confer synaptic transmission defects in the visual system were recovered. These mutationsrepresent 14 complementation groups, of which at least 9 also appear to perform functional roles outsideof the eye. Three of these complementation groups disrupt photoreceptor axonal projection, whereas theremaining complementation groups confer presynaptic defects in synaptic transmission without detectablyaltering photoreceptor structure. Mapping and complementation testing with candidate mutations revealednew alleles of the neuronal fate determinant svp and the synaptic vesicle trafficking component lap amongthe collection of mutants recovered in this screen. Given the tools available for investigation of synapticfunction in Drosophila, these mutants represent a valuable resource for future analysis of synapse develop-ment and function.

MANY of the factors responsible for axonal path- visual system (Hotta and Benzer 1969; Pak et al. 1969;Heisenberg 1971) or have been carried out under con-finding, synapse formation, and synaptic functionditions favoring the recovery of conditional allelesin metazoans were first identified in classical genetic(Suzuki 1970; Siddiqi and Benzer 1976), and thus onlyscreens carried out in Drosophila. While the genetica fraction of the genes involved in synaptic transmissionscreening approaches used to identify these factors arewere recovered from these studies. More powerful ge-powerful, they have several significant limitations. Mostnetic screening approaches in Caenorhabditis elegans thatnotably, screens for axonal pathfinding components arecircumvent some of the limitations of the Drosophilahighly labor intensive, requiring the generation of muta-system have also been successfully used to identify fac-genized lines and the systematic screening of individualtors involved in synaptic development and functionlines using antibody- or green fluorescent protein(Brenner 1974; Jorgensen and Mango 2002). How-(GFP)-based methods to identify those with altered neu-ever, the subsequent analysis of some of the mutantsronal structure (Seeger et al. 1993; Zallen et al. 1999;recovered from these screens has been compromisedParnas et al. 2001). Although somewhat less labor inten-somewhat by the difficulty in conducting electrophysio-sive, classical genetic screens for Drosophila mutantslogical analysis in C. elegans.with altered neuronal function have primarily resulted

Over the past decade, targeted mutagenesis of candi-in the identification of genes that function only in thedate genes has largely supplanted classical genetic analy-sis of neurotransmitter release mechanisms owing torapid progress in the biochemical identification of com-

1These authors contributed equally to this work. ponents thought to act in this process (Ferro-Novick2Present address: University of Washington, Box 357730, Health Sci-

and Jahn 1994; Fernandez-Chacon and Sudhof 1999;ences Bldg., K-357, Seattle, WA 98195-7730.Lin and Scheller 2000). The recent completion of the3Present address: NASA Ames Research Center, Mail Stop N261-2,

Room 104, Moffett Field, CA 94035. C. elegans, Drosophila, and mouse genome projects has4Present address: Renovis, Inc., 270 Littlefield Ave., South San Fran- further added to the list of genes that may function

cisco, CA 94080. in neurotransmitter release (Lloyd et al. 2000). These5Corresponding author: University of Washington, Box 357730, Health

approaches have led to the identification and character-Sciences Bldg., K-357, Seattle, WA 98195-7730.E-mail: [email protected] ization of components that act at many stages of synaptic

Genetics 165: 171–183 (September 2003)

172 M. C. Babcock et al.

this analysis were obtained from the Bloomington Drosophilavesicle trafficking, including vesicle fusion with the pre-Stock Center or Berkeley Drosophila Genome Project.synaptic membrane in response to calcium influx and

Generation of mutants: One- to three-day-old male fliesvesicle recycling following fusion (Fernandez-Chacon carrying an FRT element inserted at polytene segment 82Band Sudhof 1999; Lloyd et al. 2000; Richmond and were mutagenized by feeding an EMS-containing sucrose solu-

tion as described (Grigliatti 1998). Mutagenized males wereBroadie 2002). While these studies have provided in-then crossed to virgin females of one of the following geno-sight into the mechanisms of neurotransmitter release,types: y, w; EGUF/EGUF; FRT82B, GMR-hid/TM6C or y, w;a potential limitation of this approach is that it is likelyEGUF/EGUF; FRT 82B, GMR-hid/TM6,y� (both genotypes

biased in favor of those factors most readily amenable hereafter collectively designated EGUF-hid 3R). Male nonbal-to biochemical analysis. Another challenge with this ap- ancer chromosome offspring from these crosses were subjected

to an assay of phototaxis (described below).proach is that it often takes substantial time and effortMutants with sd 1-sd 15 allele designations were recovered byto generate mutations in the genes of interest and, in

placing �100 mutagenized flies with normal eye morphologyseveral cases, the resulting mutants have little or no into a countercurrent apparatus (Benzer 1967) and providingphenotype (Rosahl et al. 1993; Geppert et al. 1994; flies 15 sec to move at least half the distance of the apparatusMcMahon et al. 1996) or display phenotypes unrelated toward a light source. This phototactic selection was repeated

five times. Flies that failed to move toward the light in atto presynaptic function (Leventis et al. 2001; Razzaqleast three trials were recovered and tested again in identicalet al. 2001; Zelhof et al. 2001; Andrews et al. 2002;fashion the following day. Flies exhibiting phototactic defectsMurthy et al. 2003). on successive days were individually mated to the EGUF-hid

Recently, some of the limitations of the previous classi- 3R stock to generate a population of flies bearing the samecal genetic and biochemical approaches have been over- mutagenized chromosome and assayed for phototactic defects

as described above. Mutants exhibiting phototactic defects ascome by the development of the EGUF/hid systema population were crossed to a TM6B/TM3 stock to recover(Stowers and Schwarz 1999). This system combinesthe mutagenized chromosome over the TM6B balancer chro-

the FLP/FRT method of generating mosaic tissues in mosome. From a screen of 18,500 mutagenized flies, 171 pho-Drosophila with the GAL4/UAS system to target mitotic totactic mutants were recovered of which 15 appear to be

specifically defective in synaptic transmission on the basis ofrecombination to cells that make up the retina of thethe results of electroretinogram recordings (see below).compound eye. The presence of an eye-specific, domi-

Mutants with sd 16-sd32 allele designations were isolated bynant, proapoptotic factor eliminates all cells that doplacing up to 500 mutagenized flies into a 500-ml flask and

not bear homozygous clones of a mutated chromosome providing flies 15–20 sec to move into an adjacent 500-ml flaskarm. This system can be used to generate F1 progeny toward a fluorescent light source. Flies with normal external

eye morphology that failed phototactic selection on two suc-from a mutagenized parent that are homozygous for acessive trials were retested in identical fashion the followingmutagenized chromosome arm in the retina but hetero-day. Flies that again failed the phototaxis selection were indi-zygous elsewhere. Subsequent screening to identify fliesvidually mated to EGUF-hid 3R females. Appropriate progeny

with phototactic defects and electroretinogram alter- from this cross were subjected to electroretinogram recordingsations indicative of a defect in synaptic transmission can to identify mutants with defects in synaptic transmission as

described below. Mutants exhibiting the desired electroretino-be used to recover mutants of interest.gram characteristics were mated to y w; Sp/CyO y�; Ly/TM6In this study, we describe the preliminary results ofy� females to recover chromosomes of interest in trans toa screen for mutations mapping to the right arm ofthe TM6 y� balancer chromosome. From a screen of 24,000

chromosome 3, representing approximately one-fifth mutagenized flies, 17 mutants with presynaptic defects in syn-of the Drosophila genome, that result in presynaptic aptic transmission were recovered.

Electroretinogram analysis of mutants: Balanced stocksdefects in synaptic transmission. From this screen, 14bearing the mutations conferring nonphototactic phenotypescomplementation groups were identified, of which 11were crossed to the EGUF-hid 3R stock to generate offspringappear to specifically affect presynaptic function and 3possessing eyes homozygous for the relevant chromosome.

affect axonal pathfinding. All of these mutations have Electroretinogram recordings were carried out on these fliesbeen mapped and nearby candidate genes have been following dark adaptation as described (Pak et al. 1969).

Complementation analysis, mapping, and lethal phase analy-identified and, where possible, tested for complementa-sis of synaptic transmission mutants: Complementation analy-tion with our mutants. One of the complementationsis was carried out in two ways: First, balanced stocks of eachgroups exhibiting an aberrant axonal projection patternof the different mutants were crossed to one another in all

appears to represent the seven up gene, and one with possible combinations and offspring were scored for the pres-unaltered photoreceptor structure represents the lap ence of viable nonbalancer progeny. Mutations that failed to

complement each other regarding viability were consideredgene. This work provides a foundation for the identifi-allelic. For those mutants that produced viable offspring fromcation of novel components involved in neuronal devel-the complementation crosses, trans-heterozygous nonbalanceropment and function. offspring were subjected to an assay of phototaxis as describedabove. Mutations that failed to complement each other forthe phototaxis phenotype were further characterized by con-ducting electroretinogram (ERG) recordings to confirm theirMATERIALS AND METHODSallelic relationship.

Drosophila husbandry: All crosses were carried out at 22�, and Deficiency mapping was conducted by crossing balancedstocks of all of the synaptic transmission mutants to a collectionstocks were maintained on cornmeal agar. The stocks used in

173Synaptic Transmission Screen

of deficiency stocks, which together span most of the rightarm of chromosome 3. Mutations were tentatively assigned tothe deficiency intervals of those deficiencies that failed toproduce viable hemizygous offspring. Recombinational map-ping was carried out by crossing the synaptic transmissionmutants to a stock bearing an FRT element at polytene posi-tion 82B and the recessive markers ru1, h1, th1, st1, cu1, sr1, es,and ca1. Female offspring from this cross were then mated toa stock lacking the FRT element at 82B but bearing the samerecessive markers to identify recombinants. A total of 100male recombinants from this cross were then selected andindividually mated to the EGUF-hid 3R stock. The appropriatemale offspring from these crosses were subjected to a test ofphototaxis as described above. The map positions of mutationsresponsible for phototactic phenotypes were calculated by de-termining the average recombination frequency and standarddeviations obtained from linked markers. Mutations that pro-duced consistent results in the deficiency and recombinationalmapping exercise were assigned a localization defined by thedeficiencies that fail to rescue the recessive lethal phenotypeand in some cases were further delimited by overlapping com-plementing deficiencies.

Those mutations that produced conflicting results in thedeficiency and recombinational mapping exercise or that com-plemented all of the deficiency chromosomes were crossedagain to deficiencies mapping to the polytene regions impli-cated from recombinational mapping. Viable offspring fromthese crosses were tested for phototactic and ERG phenotypesas described above. Those mutations that complemented allof the deficiencies tested were subjected to further recombina-

Figure 1.—Flow chart diagramming the identification andtional mapping to refine the genetic map position. Recombi-characterization of photoreceptor synaptic transmission defec-national mapping was performed as described above, but onlytive mutants recovered in this work.animals with recombinational events near linked markers were

used in this analysis. Mutations were localized to the cytologicalintervals defined by deficiency mapping experiments or wereassigned a genetic map position determined from the averagerecombination frequency and standard deviations obtained horseradish peroxidase-coupled secondary antibody (Jacksonfrom linked markers if they complemented all of the available Immunoresearch, Bar Harbor, ME) in PBSTNGS for 2 hr atdeficiencies. room temperature. Following incubation with the secondary

To facilitate lethal-phase analysis, mutant chromosomes antiserum, head sections were washed three times for 5 minwere placed in trans to a balancer chromosome marked with in PBST and developed for 10 min in PBS containing 0.5%GFP. Homozygous non-GFP offspring from each stock were diaminobenzidine and 0.03% hydrogen peroxide. Head sec-collected and monitored until lethality occurred. In addition, tions immunostained for synaptotagmin were processed asthose mutants that map to deficiencies were crossed to stocks above, except that a rabbit anti-synaptotagmin antiserum was

used at 1:1000 as the primary antiserum (a gift from Troybearing the relevant deficiency chromosome in trans to a GFP-Littleton; Littleton et al. 1993) with a horseradish peroxi-marked balancer chromosome. In both analyses, non-GFPdase-coupled goat anti-rabbit secondary antibody (Jackson Im-offspring were monitored until lethality occurred.munoresearch).Candidate genes corresponding to the mutations recovered

in this analysis were tested by obtaining stocks bearing muta-tions in the candidate genes (where possible) and by per-forming complementation analysis. RESULTS

Analysis of photoreceptor axonal projection patterns:Screening for synaptic transmission mutants: To iden-Heads were prepared for immunohistochemistry by removing

the proboscis and air sacs ventral to the brain and then fixing tify mutants with defective synaptic function in the visualin PBS plus 4% paraformaldehyde for 3 hr at 4�. Fixed heads system, male flies homozygous for an FRT element atwere then rinsed 10 min in PBS plus 12% sucrose, incubated polytene position 82B were mutagenized with EMS andin 25% sucrose overnight at 4�, and then frozen in OCT

crossed to EGUF-hid 3R females (see Figure 1 for anfreezing medium (VWR) and sectioned at 10–12 �m. Headoverview of the screen and materials and methodssections were rinsed two times for 5 min each in PBST (PBS

plus 0.5% Triton X-100), blocked 1 hr at room temperature for further details). A total of 42,500 F1 offspring fromin PBSTNGS (PBST plus 5% normal goat serum), and then this cross, homozygous for the right arm of chromosomeincubated 2 hr at room temperature (or overnight at 4�) with 3 in the retina, were subjected to a test of phototaxis.a 1:50 dilution of mAb24B10 (obtained from Developmental Nonphototactic flies with normal external eye morphol-Studies Hybridoma Bank, University of Iowa; Fujita et al.

ogy were recovered and crossed to the EGUF-hid 82B1982). Following incubation with the primary antiserum, headstock to produce a population of offspring bearing thesections were washed three times for 5 min each in PBST

and incubated with a 1:300 dilution of a goat anti-mouse same mutagenized chromosome. These F2 flies were


then tested as a population to verify the original pho-totactic phenotype (sd1-15 alleles) or directly subjectedto electrophysiological analysis (sd16-32 alleles).

Mutants with defects in synaptic transmission wereidentified from among the collection of phototactic mu-tants by carrying out ERG recordings. The ERG moni-tors electrical activity in the compound eye in responseto light and consists of the sum of two components: anegative component corresponding to phototransduc-tion in the rhabdomere and a positive component re-sulting from neurotransmitter-dependent hyperpolar-ization of second-order neurons in the lamina that arepostsynaptic to the photoreceptor cells. The synapticresponse of laminal neurons results in transient upwardand downward deflections in the ERG trace upon theinitiation and cessation, respectively, of the light stimu-lus (commonly referred to as “ON” and “OFF” tran-sients, respectively; Figure 2A). The ON/OFF transientsare dependent upon release of the chemical neurotrans-mitter histamine from the photoreceptor terminal inresponse to light (Burg et al. 1993). Thus, mutantsthat are capable of light-induced depolarization of thephotoreceptor cells but are defective in synaptic trans-mission will specifically lack ON/OFF transients but willretain the sustained phototransduction component ofthe ERG. Further, because the ey-GAL4 line used to drivemitotic recombination in the EGUF/hid system is notexpressed in the lamina, recessive mutations conferringON/OFF transient defects recovered from our screenwill be specific to the presynaptic photoreceptor neu-

Figure 2.—Electroretinogram recordings of selected mu-rons.tants recovered in this analysis. All ERG recordings were car-This analysis led to the identification of 32 mutationsried out on animals bearing homozygous eye clones of thethat preferentially affect the ON/OFF transients of the relevant chromosome obtained from crosses to the EGUF/hid

ERG. These synaptic transmission defective mutants 3R stock. (A) ERG recording from a wild-type control animalwere given the allele designations sd1-32. Many of the bearing the parental chromosome used in genetic analysis

showing the ON and OFF transients (arrows) and sustainedmutants recovered in this analysis lack ON/OFF tran-component (arrowhead) of the ERG. Light pulses are desig-sients under all conditions tested, as exemplified bynated by horizontal black bars shown below the ERG re-

mutant sd3 (Figure 2B; Table 1). However, in contrast cordings. (B) ERG recording from the sd3 mutant recoveredto wild-type flies, which maintain robust ON/OFF tran- in this analysis. Note the absence of the ON and OFF transients

(arrows) and retention of photoreceptor cell depolarizationsients during repetitive light stimulation (Figure 2C),(arrowhead) in this mutant. (C) ERG recording from a wild-six mutants in our collection exhibit ON and/or OFFtype animal stimulated repetitively with �1-sec duration each.transients that are progressively lost during repetitive (D) ERG recording from the sd 15 mutant stimulated with inter-

light stimulation when measured after a period of dark mittent light and dark pulses of �1-sec duration each. Thisadaptation, as exemplified by mutant sd15 (Figure 2D; mutant lacks an ON transient, but retains an OFF transient

that is lost gradually upon repetitive stimulation. (E) ERGTable 1). Finally, four mutants in our collection retainrecording from the sd 10 mutant after intermittent light andON transients but exhibit a progressive loss of the OFFdark pulses of 0.5 sec each for 10 sec, revealing the absence

transient upon repetitive light stimulus, as shown for of only the OFF transient.the sd10 mutant (Figure 2E; Table 1).

Complementation analysis and mapping of synaptictransmission mutants: The 32 synaptic transmission mu-tants recovered from this screen were crossed to one carried out to identify mutations that fail to complementanother in all possible combinations to identify allelic each other for a recessive lethal phenotype. However,relationships. Because each of the chromosomes recov- offspring from those crosses that produced viable trans-ered from this screen confers a recessive lethal pheno- heterozygotes were also tested for phototactic and ERGtype that may derive from the mutation responsible phenotypes. Results of this analysis indicate that these

32 mutants represent 14 different complementationfor the synaptic transmission defect, crosses were first


TABLE 1

Electroretinogram phenotypes

Defective ON transient andactivity-dependent loss of Activity-dependent loss

Defective ON/OFF transients OFF transient of OFF transient only

sd 1, sd2, sd3, sd4, sd5, sd 6, sd7, sd 8, sd9, sd 11, sd 12, sd 13, sd 15, sd 19 sd 10, sd23, sd26, sd32

sd 14, sd 16, sd 17, sd 18, sd20, sd21, sd22, sd 14, sd25, sd27,sd28, sd29, sd30, sd31

groups with between one and six alleles each (Table 2). have presynaptic defects in synaptic transmission, thisphenotype could arise either from defective neurotrans-The large fraction of complementation groups with only

a single allele indicates that this screen was not saturat- mitter release from the photoreceptor neurons or fromfailure of the photoreceptor neurons to properly forming despite the high dose of mutagen used and the

large number of mutagenized animals tested. Of the 8 synaptic connections with their targets in the lamina.To distinguish between these possibilities, the photore-complementation groups with more than one allele

each, 6 produce trans-heterozygous lethal phenotypes, ceptor axonal projection patterns of the synaptic trans-mission mutants were analyzed in head sections usingand 2 appear to be viable as trans-heterozygotes. While

many of the allelic combinations of complementation the photoreceptor-specific monoclonal antibody mAb24-B10. Results of this analysis using a parental control linegroup 1 are lethal as trans-heterozygotes, several allelic

combinations result in viable offspring that are blind are shown in Figure 3A. Identical experiments carriedout with mutations representing the 14 complementa-and lack ON/OFF transients (Table 2). These viable

combinations indicate that at least several of the completion groups recovered from our screen revealed 3 com-plementation groups that exhibit a grossly abnormalmentation group 1 mutations represent hypomorphic

alleles. Although no definitive conclusion can be drawnfrom the 8 complementation groups with a single alleleeach, recombinational and deficiency mapping (see be- TABLE 2low) suggests that at least 2 of these complementation

Complementation analysisgroups define genes that are essential for viability.The synaptic transmission mutants were mapped in

Complementation Viable astwo ways. All of the mutants were first crossed to agroups Mutations trans -heterozygote?

collection of stocks, each bearing a deficiency mapping1 sd 8, sd 13, sd 14, sd 17, Noato the right arm of chromosome 3. This collection of

sd24, sd31stocks, in aggregate, carries deletions of �85% of the2 sd 12, sd 16, sd20, sd29 Nosequences on 3R. The deficiency stocks were used to3 sd 10, sd23, sd26, sd32 Nolocalize mutations responsible for recessive lethal phe-4 sd 1, sd 11, sd21 Yes

notypes. Second, all of the mutations were mapped by 5 sd4, sd 6, sd 18 Norecombination using a chromosome containing multi- 6 sd2, sd 15 Yesple recessive markers. Recombinational mapping was 7 sd22, sd27 No

8 sd7, sd25 Nocarried out to verify that mutations responsible for reces-9 sd28 NAsive lethal phenotypes map close to, and thus likely

10 sd30 NAbrepresent, the synaptic transmission mutations. Synaptic

11 sd 19 NAtransmission mutations that complemented all defi-12 sd3 NAb,c

ciencies in the collection or produced conflicting results 13 sd 5 NAc

in the deficiency and recombinational mapping analyses 14 sd9 NA(indicating that the mutation identified from deficiency

NA, not applicable.mapping is incidental) were again crossed to stocks bear- a Three allele combinations in this complementation grouping deficiency chromosomes in the regions implicated are lethal as trans-heterozygotes, specifically, sd 8/sd 17, sd 14/from recombinational mapping. Offspring from this sd 17, and sd 17/sd24. All other combinations produce viable trans-

heterozygotes that are blind and exhibit ON/OFF transientcross were analyzed for phototactic and ERG pheno-defects, except mutant sd31, which exhibits an ON/OFF tran-types to investigate the possibility that these mutantssient phenotype only in trans to sd 17.are viable or semiviable as hemizygotes. Results of this b Rare homozygous adult escapers that lack ON and OFF

analysis are summarized in Table 3. transients are observed.Phenotypic analysis of synaptic transmission mutants: c These mutations are lethal in combination with deficien-

cies that map to the same region.While all of the mutants recovered from our screen


TA

BL

E3

Mei

otic

and

defi

cien

cym

appi

ngre

sult

s

Ove

rlap

pin

gco

mpl

emen

tin

gC

ompl

emen

tati

onR

ecom

bin

atio

nal

map

posi

tion

Non

com

plem

enti

ng

defi

cien

cies

ade

fici

enci

esb

grou

p(c

orre

spon

din

gpo

lyte

ne

loca

lizat

ion

)(d

elet

ion

brea

kpoi

nts

)(d

elet

ion

brea

kpoi

nts

)M

appo

siti

onc

186

.5�

7.4

(94–

98)

Df(

3R)X

Sd,e

(096

A01

-07;

096A

21-2

5),

Df(

3R)9

6B(0

96A

21;0

96B

08-1

0)96

A01

-96A

18D

f(3R

)crb

87-5

d,e

(095

F07;

096A

17-1

8)2

82.4

�3.

8(9

4–95

)D

f(3R

)345

0d

(098

E03

;099

A06

-08)

Df(

3R)D

r-rv1

(099

A01

-02;

099B

06-1

1)98

E03

-99A

023

81�

0.4

(95–

95)

Df(

3R)m

bc-R

1d(0

95A

05-0

7;09

5D06

-11)

Df(

3R)m

bc-3

0(0

95A

05-0

7;09

5C10

-11)

95C

10-9

5D11

453

.6�

13.9

(69–

92)

Df(

3R)M

-Kx1

e(0

86C

01;0

87B

01-0

5),

Df(

3R)T

-32

(086

E02

-04;

087C

06-0

7)86

D01

-86D

04D

f(3R

)M86

De

(86D

01-0

2;86

D04

)5

63�

0.1

(91)

Df(

3R)D

G4d

(090

D02

-04;

090F

03-0

6),

Df(

3R)C

ha7

(90F

01-0

2;91

F5)

90D

02-9

0F02

Df(

3R)P

14d

(090

C02

-D01

;091

A01

-02)

647

.6�

0.6

(77–

85)

Non

e82

Bf -8

57

54.2

�3.

5(8

6–89

)D

f(3R

)M-K

x1d

(086

C01

;087

B01

-05)

,86

E02

-87B

05D

f(3R

)T-3

2d

(086

E02

-04;

087C

06-0

7)8

49�

3.1

(76–

87)

Non

e82

Bf -8

79

52.8

�3.

5(8

5–88

)N

one

85-8

810

49.4

�0.

785

(85–

86)

Df(

3R)b

y62

d(0

85D

11-1

4;08

5F06

),85

D11

-85F

01D

f(3R

)by1

0d

(085

D08

-12;

085E

07-F

01)

1110

1.8

�3.

4(9

9–10

0)N

one

99-1

0012

45.6

�10

.4(6

8–88

)D

f(3R

)dsx

2Md

(084

C01

-03;

084E

01),

84C

03-8

4D02

Df(

3R)A

ntp1

7d

(084

B01

-02;

084D

11-1

2or

A06

,D14

),D

f(3R

)Scx

4d

(084

B03

;084

D01

-02)

,D

f(3R

)An

tp1d

(084

B03

;084

D01

-02)

1346

.4�

2.9

(71–

86)

Df(

3R)p

40d

(084

E08

-09;

085B

06),

Df(

3R)p

13(0

84F0

2;08

5B01

)84

E12

-84F

02D

f(3R

)CA

1d(0

84E

12-1

3;08

5A06

-11)

,D

f(3R

)CA

3d(0

84F0

2;08

5A05

-07)

1468

.3�

1.1

(92)

Non

e92

aN

onco

mpl

emen

tin

gde

fici

enci

esar

eth

ose

that

fail

toco

mpl

emen

ta

rece

ssiv

ele

thal

phen

otyp

eor

aph

otot

acti

cph

enot

ype

and

map

toth

ere

gion

impl

icat

edby

reco

mbi

nat

ion

alm

appi

ng.

bD

efici

enci

esth

atco

mpl

emen

tth

esy

nap

tic

tran

smis

sion

mut

atio

ns

but

over

lap

wit

hn

onco

mpl

emen

tin

gde

fici

enci

es.

cPo

lyte

ne

map

posi

tion

sde

fin

edby

defi

cien

cyan

d/or

reco

mbi

nat

ion

alm

appi

ng

(see

text

).Po

lyte

ne

map

posi

tion

sin

ferr

edso

lely

from

reco

mbi

nat

ion

alm

appi

ng

are

unde

rlin

ed.

dD

efici

ency

that

fails

toco

mpl

emen

tre

cess

ive

leth

alph

enot

ype.

eD

efici

ency

that

fails

toco

mpl

emen

tph

otot

acti

can

dE

RG

phen

otyp

es.

fL

eft

limit

dete

rmin

edby

FRT

loca

tion

atpo

lyte

ne

posi

tion

82B

.


Figure 3.—Photoreceptor axon and opticlobe neuropil labeling of homozygous eyes ofsynaptic transmission mutants. (A–I) Cryostatsections of Drosophila heads homozygous inthe eye for the indicated genotypes labeledwith the photoreceptor axon marker mAb24-B10 (A–E) or the synaptic vesicle-specific anti-body synaptotagmin that exclusively labelsneuropil (F–I). (A) Parental control showingthe ordered array of R1-6 photoreceptor axonlabeling in the lamina and R7-8 labeling in themedulla. (B) Complementation group 1 (sd31)exhibiting an axon projection pattern essen-tially indistinguishable from control. (C) Com-plementation group 2 (sd 16), (D) complemen-tation group 7 (sd22), and (E) complementationgroup 9 (sd28) all show grossly abnormal photo-receptor projection patterns. (F) Parental con-trol showing the normal optic neuropil pat-tern. (G) Complementation group 2 (sd 16),(H) complementation group 7 (sd22), and (I)complementation group 9 (sd28) all showgrossly abnormal laminas and medullas. In Gand H the medulla is �90� from its normalposition relative to the lamina. In I the lobulaand lobula plate are �90� from their normalposition relative to the rhabdomeres, sug-gesting that in all three cases some rotationalevent of the medulla, lobula, and lobula platehas occurred abnormally. lam, lamina; med,medulla; lob, lobula; lob pl, lobula plate.

axonal projection pattern (Figure 3, C–E). Additionally, that is indistinguishable from that of the control lineused in this analysis (Figure 3B; data not shown). How-development of the lamina, which requires innervation

by retinal neurons, is partially disrupted in mutant sd16 ever, it remains possible that subtle defects in axonpathfinding/synaptogenesis that are below the level ofand sd22 (corresponding to complementation groups 2

and 7, respectively) and is completely disrupted in mu- detection of this analysis are present in these mutants.To further characterize the three complementationtant sd28 (corresponding to complementation group 9).

Similar axonal and laminal disruption was observed in groups exhibiting altered photoreceptor axonal path-finding/synaptogenesis, head sections from representa-all other alleles corresponding to complementation

groups 2 and 7, confirming the role of these comple- tive mutants corresponding to these complementationgroups were stained with an antiserum to the synapticmentation groups in photoreceptor axon pathfinding

and/or synaptogenesis. All of the remaining comple- vesicle protein synaptotagmin. This antiserum specifi-cally labels the neuropil in the optic lobe of controlmentation groups exhibit an axonal projection pattern


TABLE 4

Lethal-phase analysis of synaptic transmission mutants

Complementation Hemizygous Homozygousgroup Allele lethal phasea,b lethal phaseb

1 sd 8 First Secondsd 13 Viable adults Pharate adultssd 14 First Secondsd 17 First/second Secondsd24 Viable adults Viable adultssd31 Viable adults First

2 sd 12 Embryo Embryosd 16 Embryo/first Embryosd20 Embryo Embryosd29 Embryo Embryo

3 sd 10 Pharate adults Pharate adultssd23 Pupae Firstsd26 Pupae Firstsd32 Pupae Third

5 sd 4 Second Second/thirdsd 6 Second Secondsd 18 Second Third instar

7 sd22 Embryo/first Embryo/firstsd27 Embryo Embryo

8 sd7 NA Firstsd25 NA Embryo

9 sd28 NA Second-third10 sd30 Pupae Pupae/adult (rare escapers)11 sd 19 NA First-second12 sd3 First Pupae/adult (rare escapers)13 sd 5 Third/pupae Third/pupae14 sd9 NA First

NA, not applicable.a Lethal phase in trans to a noncomplementing deficiency.b First, second, and third indicate first, second, and third instar larval lethal phases, respectively.

flies (Figure 3F). In contrast to the staining observed phase may correspond to or be influenced by incidentalmutations residing on the same chromosome. Such anin control flies, the representatives of complementation

groups 2, 7, and 9 showed an altered synaptotagmin occurrence is likely indicated for mutations conferring amore severe phenotype as a homozygote than as a hemizy-labeling pattern (Figure 3, G–I). Only partial laminal

staining was observed in the mutants corresponding gote (as seen for mutations sd13, sd23, sd26, sd31, and sd32).Results of this analysis are summarized in Table 4.to complementation groups 2 and 7, and the normal

ordered morphology of the medulla is disrupted. Fur-thermore, while the location of the lamina relative to

DISCUSSIONthe overlying photoreceptors is normal in these twocomplementation groups, the remainder of the optic We have conducted a screen using the EGUF-hid sys-

tem to identify recessive mutations that disrupt synapticlobe is misoriented �90 degrees relative to the lamina.The morphology of the lamina/medulla region of com- transmission in the Drosophila visual system. This system

generates flies bearing eyes composed exclusively ofplementation group 9 is even more irregular than that incomplementation groups 2 and 7 and exhibits a similar homozygous clones for a selected chromosome arm and

provides an efficient means for conducting F1 screensdisruption of the orientation of the medulla with respectto the lamina. for mutations affecting neuronal structure and func-

tion, even if these mutations affect genes that are essen-The lethal phase of synaptic transmission mutants wasdetermined by placing mutations in trans to a noncom- tial for adult viability. In this study we screened 42,500

mutagenized flies and recovered 32 mutants represent-plementing deficiency and monitoring the stage atwhich lethality occurred. For those mutants that do not ing 14 complementation groups that preferentially af-

fect the ON/OFF transient component of the ERG.map to a deficiency, lethal-phase analysis was conductedby monitoring homozygotes. A risk associated with le- Most of the mutations recovered also confer a recessive

lethal phenotype, suggesting that these genes may playthal-phase analysis of homozygotes is that the lethal


a general role in synaptic transmission in the nervous molecular nature of the lapKG06751 allele (which bears atransposon insertion early in the coding sequence of thesystem. However, at least 4 complementation groups (4,first exon of lap) indicates that this allele may represent a6, 10, and 12) produce viable adults in trans-heterozy-lap null. Further complementation testing excluded thegous or hemizygous configuration (Table 2). These mu-syntaxin1A gene as a candidate of complementationtations may reside in genes that function only in thegroup 3. However, AP1�, rab7, syntaxin18, and sec10visual system or may affect genes that function moremutants have not previously been identified, and thusbroadly but only detectably affect the visual system. Forfurther work will be required to test whether mutationsthe remaining complementation groups, it is not possi-in these genes underlie the phenotypes we have docu-ble at this time to conclude definitively whether themented. While most of the candidate genes listed ingenes are essential for viability.Figure 4 correspond to axonal pathfinding componentsThe largest category of mutations in our collectionand components of the vesicle trafficking pathway, theconfers ON/OFF transient defects in the visual systemrecent identification of milton, a Drosophila gene in-without detectably altering the normal pattern of photo-volved in the transport of mitochondria to synaptic ter-receptor axonal projection to targets in the lamina andminals, from an EGUF/hid screen demonstrates that ourmedulla. Because the EGUF-hid system produces homo-mutants could represent a much broader collection ofzygous clones only in the photoreceptor cells, whereasgenes than those displayed in Figure 4 (Stowers et al.cells in the optic lobe remain heterozygous, the recessive2002).phenotypes induced by this collection of mutations must

A subset of the mutants recovered in this analysisderive from presynaptic defects in synaptic transmission.retains some residual synaptic function that is lost uponThe fact that these mutants exhibit a substantial sus-repetitive stimulation. Similar phenotypes have beentained component of the ERG indicates that the molecu-shown to result from mutations in genes that function inlar defect in these mutants derives from a failure insynaptic vesicle priming or recycling, including Shibire,the release of the chemical neurotransmitter histamineNSF1, and Endophilin1A (Salkoff and Kelly 1978;from the photoreceptor nerve terminal at a step down-Poodry and Edgar 1979; Koenig and Ikeda 1989;stream of light-induced depolarization. Previous analy-Kawasaki et al. 1998; Rikhy et al. 2002). This phenotypeses have shown that mutants with defects in histamineis thought to arise from a reduced ability to maintainbiosynthesis and synaptic vesicle fusion and recyclinga readily releasable pool of synaptic vesicles, which isbehave identically to this category of mutant (Burg etrevealed only upon rapid repetitive stimulation. Thus,al. 1993; Littleton et al. 1998; Stowers and Schwarzthe complementation groups displaying these pheno-1999; Sanyal et al. 2001). Thus, genes involved in thetypes may encode products that participate in synaptic

reuptake and packaging of histamine into synaptic vesi-vesicle priming or recycling. Although it remains un-

cles or the targeting, fusion, or recycling of synaptic clear why several mutants in this category exhibit activity-vesicles at release sites are likely represented among dependent loss of only the OFF transient, it is worththese mutants. noting that a similar phenotype was reported for the

A number of our mutations that show normal axonal synaptic vesicle recycling mutant, endophilin1A, demon-morphology map to regions encompassing Drosophila strating that this characteristic is not incompatible withgenes implicated in synaptic vesicle trafficking and thus a role in synaptic vesicle trafficking.these genes represent candidate genes to those muta- The other major category of mutant recovered in ourtions recovered in our screen (Figure 4; black text). screen consists of those that lack ON/OFF transientsFor example, the sd3 mutant maps to a deficiency that because the photoreceptor neurons fail to properly pro-removes the genes encoding the synaptic vesicle-recy- ject axons to their target cells in the optic lobe. Thecling component LAP and the calcium-binding protein photoreceptor cells in this collection of mutants (com-synaptotagmin IV. Likewise, complementation group 3 plementation groups 2, 7, and 9) fail to choose a singlemutants are lethal in combination with deficiencies that side of the medulla to project down and subsequentlyremove the genes encoding the synaptic vesicle recy- extend down both sides. A similar phenotype has beencling adaptor protein AP1�, the trafficking proteins observed for the irreC-rst gene (Schneider et al. 1995).RAB7 and SEC10, and the t-SNAREs syntaxin1A and The proper extension of the R8 axons down one sidesyntaxin 18. Complementation analysis revealed two dif- of the medulla is needed for the reorientation of theferent alleles of the lap gene, lapKG06751 and lap1, which medullar neurons relative to the lamina (Wolff et al.failed to complement and weakly complemented the 1997; Clandinin and Zipursky 2002). Thus, the orien-recessive lethal phenotype of sd3, respectively, indicating tation of the lamina and medulla that is established inthat sd3 is a new allele of lap. The weak complementation third instar larvae is incorrectly retained in the adultof lap1 with sd3 and occasional appearance of homozy- (compare Figure 3F to Figure 3, G–I). For complemen-gous sd3 and lap1 adult escapers likely reflects the fact tation groups 2 and 7, the photoreceptor R8 axons ap-that these mutations are hypomorphic alleles of lap. By pear to have extended properly through the laminal

precursor cells, as this extension is believed necessarycontrast, the lack of homozygous viability and severe


Fig

ure

4.—

Map

posi

tion

sof

syn

apti

ctr

ansm

issi

onm

utan

tsid

enti

fied

inth

isan

alys

isan

dca

ndi

date

gen

esre

sidi

ng

onth

esa

me

chro

mos

ome

arm

.Syn

apti

ctr

ansm

issi

onm

utan

tsre

cove

red

inth

isst

udy

are

depi

cted

abov

eth

ech

rom

osom

ear

man

dca

ndi

date

gen

esar

esh

own

belo

w.

Th

eFR

Tel

emen

tus

edin

this

anal

ysis

isde

pict

edas

abl

ack

tria

ngl

en

ear

the

cen

trom

ere.

Th

em

appo

siti

ons

(in

poly

ten

eun

its)

ofth

eco

mpl

emen

tati

ongr

oups

reco

vere

din

this

wor

kar

ede

sign

ated

.Th

ose

com

plem

enta

tion

grou

psth

ath

ave

been

loca

lized

byre

com

bin

atio

non

lyar

eun

derl

ined

.Com

plem

enta

tion

grou

psth

atdo

not

affe

ctsy

nap

tic

stru

ctur

ear

ede

sign

ated

inbl

ack.

Com

plem

enta

tion

grou

psde

sign

ated

inpu

rple

repr

esen

tth

ose

con

ferr

ing

axon

alpa

thfi

ndi

ng

phen

otyp

es.C

ompl

emen

tati

ongr

oups

desi

gnat

edin

red

repr

esen

tth

ose

that

are

viab

leas

tran

s-h

eter

ozyg

otes

orh

emiz

ygot

es.C

andi

date

gen

esen

codi

ng

com

pon

ents

ofth

en

euro

tran

smit

ter

rele

ase

appa

ratu

sar

ede

sign

ated

inbl

ack.

Can

dida

tege

nes

invo

lved

inax

onal

path

fin

din

gar

ede

sign

ated

inpu

rple

.C

andi

date

gen

esth

atar

en

otes

sen

tial

for

viab

ility

are

show

nin

red.


to induce cell division in the lamina. In contrast, com- type (Miklos and Rubin 1996), we conservatively esti-mate that an average of at least 10 alleles per essentialplementation group 9 mutants appear to lack laminalgene should have been recovered. However, the largestneurons, indicating that R8 axon extension failed priorcomplementation group in our collection consists ofto reaching the lamina.only six alleles, and 6 of the 15 complementation groupsSeveral of the mutants that show axon projection de-recovered are represented by only one allele. Thesefects also map to regions containing candidate genesfindings indicate that our screen did not reach satura-implicated in axonal pathfinding (Figure 4; those high-tion, despite the large number of mutagenized animalslighted in purple). For example, complementation groupanalyzed.2 alleles map to a deficiency that removes the WASp,

The lack of saturation in our screen may derive fromDoa, Apc, and possibly the Huntingtin genes. These genesseveral different sources. Because phototactic selectionhave been implicated in sensory organ developmentwas conducted with a population of flies, the majorityand neuronal maintenance (Yun et al. 1994; Ahmed etof which were probably not blind, many blind mutantsal. 1998; Dragatsis et al. 2000; Ben-Yaacov et al. 2001).may have been swept toward the light source as part ofComplementation groups 7 and 9 map to a region thatthe “herd.” Thus, many mutants may have been lostencompasses the prospero and svp genes, which arebecause they behaved like flies with normal vision. Alter-known to be involved in axonogenesis and neurogenesisnatively, the lack of saturation observed may derive from(Mlodzik et al. 1990; Vaessin et al. 1991; Hu et al. 1998).a selection bias favoring the recovery of only particularLocalization of complementation group 9 is based solelyalleles of genes that function in photoreceptor develop-on meiotic mapping criteria, as deficiencies removingment and function. For example, null mutations of thethe sd28 allele were not identified. Thus, complementa-syntaxin1A gene fail to support retinal developmenttion group 9 spans a large region of chromosome 3R(Stowers and Schwarz 1999), and thus such allelesand many additional candidate genes exist for this com-would not have been recovered from our screen givenplementation group. Complementation analysis withour requirement that mutants exhibit normal externalprospero mutations excluded this gene as a candidate ofeye morphology. By contrast, hypomorphic alleles ofcomplementation groups 7 and 9. However, both allelesgenes required for cell viability may not reduce synapticof complementation group 7 fail to complement thefunction sufficiently to allow their recovery from a pho-recessive lethal phenotype of two different svp mutationstotaxis screen. Thus, only a restricted subset of hypomor-(svp1 and svp07842), indicating that complementationphic alleles of such genes would be expected from ourgroup 7 represents the svp gene. svp encodes a steroidscreen. While selection bias cannot be circumvented,receptor involved in photoreceptor cell fate specifica-conducting the EGUF/hid screen under conditions fa-

tion. Loss of svp function results in failure to form R3/4voring the recovery of conditional alleles would likely

and R1/6 photoreceptors with the corresponding cells extend the range of genes that could be obtained. Tem-defaulting to an R7 fate (Mlodzik et al. 1990). The perature-sensitive alleles of the shibire and NSF1 genesabsence of R3/4 and R1/6 photoreceptors or the over- result in blindness and a lack of ON/OFF transientabundance of R7 photoreceptors may be responsible for phenotypes at restrictive temperature when they arethe abnormal axonal projections seen in these mutants. made homozygous in the retina using the EGUF-hid sys-While further complementation analysis has excluded tem (our unpublished results), supporting the validitythe WASp, Doa, and Apc genes as complementation of such an approach.group 2 candidates, the Huntingtin gene remains a can- In summary, we have identified a collection of mu-didate that will be investigated in future work. tants that display presynaptic defects in synaptic trans-

While many of the mutants recovered in this analysis mission in the Drosophila visual system and have placedmap to genomic regions bearing candidate genes, a the mutants into two broad categories: those with defec-number of genes on chromosome 3R that are known tive photoreceptor axonal projection patterns and thosefrom previous work to participate in photoreceptor de- with apparently normal photoreceptor structure. Forvelopment and function were not recovered in this anal- those mutants that exhibit normal axonal structure, fur-ysis. For example, we did not recover mutations in en- ther structure/function studies using more powerfuldophilin1A, syntaxin1A, or gyc� despite the fact that electrophysiological and ultrastructural approaches willprevious work has shown that mutations in these genes be used to better define their roles in synaptic transmis-can result in ERG phenotypes like those obtained in sion. Additional mapping to further narrow the regionsour screen (Littleton et al. 1998; Gibbs et al. 2001; containing these genes, coupled with candidate geneRikhy et al. 2002). Given the frequency of obtaining complementation testing, sequencing, and transgenicrecessive lethal mutations at the dosage of mutagen rescue experiments, will facilitate the molecular identi-used in our analysis (Grigliatti 1998) and current fication of genes responsible for these phenotypes. Weestimates indicating that approximately one-third of the anticipate that identification of these genes will contrib-�3000 genes residing on the right arm of chromosome ute important insights into the molecular mechanisms

of synaptic development and function.3 can be mutated to produce a recessive lethal pheno-


Campos et al., 2001 Drosophila Amphiphysin is a post-synapticWe thank the Berkeley Drosophila Genome Project and Bloomingtonprotein required for normal locomotion but not endocytosis.Drosophila Stock Center for fly stocks used in this work. We also thankTraffic 2: 839–850.all members of the Pallanck lab for critical comments on the manu-

Lin, R. C., and R. H. Scheller, 2000 Mechanisms of synaptic vesiclescript. Finally, special thanks go to Johnny Palka for advice and assis-exocytosis. Annu. Rev. Cell Dev. Biol. 16: 19–49.

tance with electrophysiological experiments. This work was supported Littleton, J. T., H. J. Bellen and M. S. Perin, 1993 Expressionby a National Science Foundation CAREER award to L.J.P. and a of synaptotagmin in Drosophila reveals transport and localizationPublic Health Service Fellowship (1 F32) to R.S.S. (NS10561-01). of synaptic vesicles to the synapse. Development 118: 1077–1088.R.S.S. was also supported as a Research Associate of the Howard Littleton, J. T., E. R. Chapman, R. Kreber, M. B. Garment, S. D.

Carlson et al., 1998 Temperature-sensitive paralytic mutationsHughes Medical Institute.demonstrate that synaptic exocytosis requires SNARE complexassembly and disassembly. Neuron 21: 401–413.

Lloyd, T. E., P. Verstreken, E. J. Ostrin, A. Phillippi, O. Licht-arge et al., 2000 A genome-wide search for synaptic vesicle cycleLITERATURE CITEDproteins in Drosophila. Neuron 26: 45–50.

McMahon, H. T., V. Y. Bolshakov, R. Janz, R. E. Hammer, S. A.Ahmed, Y., S. Hayashi, A. Levine and E. Wieschaus, 1998 Regula-tion of armadillo by a Drosophila APC inhibits neuronal apoptosis Siegelbaum et al., 1996 Synaptophysin, a major synaptic vesicle

protein, is not essential for neurotransmitter release. Proc. Natl.during retinal development. Cell 93: 1171–1182.Andrews, H. K., Y. Q. Zhang, N. Trotta and K. Broadie, 2002 Acad. Sci. USA 93: 4760–4764.

Miklos, G. L., and G. M. Rubin, 1996 The role of the genomeDrosophila Sec10 is required for hormone secretion but notgeneral exocytosis or neurotransmission. Traffic 3: 906–921. project in determining gene function: insights from model organ-

isms. Cell 86: 521–529.Ben-Yaacov, S., R. Le Borgne, I. Abramson, F. Schweisguth andE. D. Schejter, 2001 Wasp, the Drosophila Wiskott-Aldrich Mlodzik, M., Y. Hiromi, U. Weber, C. S. Goodman and G. M. Rubin,

1990 The Drosophila seven-up gene, a member of the steroidsyndrome gene homologue, is required for cell fate decisionsmediated by Notch signaling. J. Cell Biol. 152: 1–13. receptor gene superfamily, controls photoreceptor cell fates. Cell

60: 211–224.Benzer, S., 1967 Behavioral mutants of Drosophila isolated by coun-tercurrent distribution. Proc. Natl. Acad. Sci. USA 58: 1112–1119. Murthy, M., D. Garza, R. H. Scheller and T. L. Schwarz, 2003

Mutations in the exocyst component sec5 disrupt neuronal mem-Brenner, S., 1974 The genetics of Caenorhabditis elegans . Genetics77: 71–94. brane traffic, but neurotransmitter release persists. Neuron 37:

433–447.Burg, M. G., P. V. Sarthy, G. Koliantz and W. L. Pak, 1993 Geneticand molecular identification of a Drosophila histidine decarbox- Pak, W. L., J. Grossfield and N. V. White, 1969 Nonphototactic

mutants in a study of vision of Drosophila. Nature 222: 351–354.ylase gene required in photoreceptor transmitter synthesis.EMBO J. 12: 911–919. Parnas, D., A. P. Haghighi, R. D. Fetter, S. W. Kim and C. S.

Goodman, 2001 Regulation of postsynaptic structure and pro-Clandinin, T. R., and S. L. Zipursky, 2002 Making connections inthe fly visual system. Neuron 35: 827–841. tein localization by the Rho-type guanine nucleotide exchange

factor dPix. Neuron 32: 415–424.Dragatsis, I., M. S. Levine and S. Zeitlin, 2000 Inactivation ofHdh in the brain and testis results in progressive neurodegenera- Poodry, C. A., and L. Edgar, 1979 Reversible alteration in the

neuromuscular junctions of Drosophila melanogaster bearing ation and sterility in mice. Nat. Genet. 26: 300–306.Fernandez-Chacon, R., and T. C. Sudhof, 1999 Genetics of synap- temperature-sensitive mutation, shibire. J. Cell Biol. 81: 520–527.

Razzaq, A., I. M. Robinson, H. T. McMahon, J. N. Skepper, Y. Sutic vesicle function: toward the complete functional anatomy ofan organelle. Annu. Rev. Physiol. 61: 753–776. et al., 2001 Amphiphysin is necessary for organization of the

excitation-contraction coupling machinery of muscles, but notFerro-Novick, S., and R. Jahn, 1994 Vesicle fusion from yeast toman. Nature 370: 191–193. for synaptic vesicle endocytosis in Drosophila. Genes Dev. 15:

2967–2979.Fujita, S. C., S. L. Zipursky, S. Benzer, A. Ferrus and S. L. Shot-well, 1982 Monoclonal antibodies against the Drosophila ner- Richmond, J. E., and K. S. Broadie, 2002 The synaptic vesicle cycle:

exocytosis and endocytosis in Drosophila and C. elegans. Curr.vous system. Proc. Natl. Acad. Sci. USA 79: 7929–7933.Geppert, M., V. Y. Bolshakov, S. A. Siegelbaum, K. Takei, P. De Opin. Neurobiol. 12: 499–507.

Rikhy, R., V. Kumar, R. Mittal and K. S. Krishnan, 2002 Endophi-Camilli et al., 1994 The role of Rab3A in neurotransmitterrelease. Nature 369: 493–497. lin is critically required for synapse formation and function in

Drosophila melanogaster. J. Neurosci. 22: 7478–7484.Gibbs, S. M., A. Becker, R. W. Hardy and J. W. Truman, 2001 Solu-ble guanylate cyclase is required during development for visual Rosahl, T. W., M. Geppert, D. Spillane, J. Herz, R. E. Hammer et

al., 1993 Short-term synaptic plasticity is altered in mice lackingsystem function in Drosophila. J. Neurosci. 21: 7705–7714.Grigliatti, T. A., 1998 Mutagenesis, pp. 55–64 in Drosophila: A synapsin I. Cell 75: 661–670.

Salkoff, L., and L. Kelly, 1978 Temperature-induced seizure andPractical Approach, edited by B. D. Roberts. Oxford UniversityPress, New York. frequency-dependent neuromuscular block in a ts mutant of Dro-

sophila. Nature 273: 156–158.Heisenberg, M., 1971 Separation of receptor and lamina potentialsin the electroretinogram of normal and mutant Drosophila. J. Sanyal, S., L. A. Tolar, L. Pallanck and K. S. Krishnan, 2001 Ge-

netic interaction between shibire and comatose mutations inExp. Biol. 55: 85–100.Hotta, Y., and S. Benzer, 1969 Abnormal electroretinograms in Drosophila suggest a role for snap-receptor complex assembly

and disassembly for maintenance of synaptic vesicle cycling. Neu-visual mutants of Drosophila. Nature 222: 354–356.Hu, S., M. Sonnenfeld, S. Stahl and S. T. Crews, 1998 Midline rosci. Lett. 311: 21–24.

Schneider, T., C. Reiter, E. Eule, B. Bader, B. Lichte et al., 1995Fasciclin: a Drosophila Fasciclin-I-related membrane protein lo-calized to the CNS midline cells and trachea. J. Neurobiol. 35: Restricted expression of the irreC-rst protein is required for nor-

mal axonal projections of columnar visual neurons. Neuron 15:77–93.Jorgensen, E. M., and S. E. Mango, 2002 The art and design of 259–271.

Seeger, M., G. Tear, D. Ferres-Marco and C. S. Goodman, 1993genetic screens: Caenorhabditis elegans. Nat. Rev. Genet. 3: 356–369. Mutations affecting growth cone guidance in Drosophila: genes

necessary for guidance toward or away from the midline. NeuronKawasaki, F., A. M. Mattiuz and R. W. Ordway, 1998 Synapticphysiology and ultrastructure in comatose mutants define an in 10: 409–426.

Siddiqi, O., and S. Benzer, 1976 Neurophysiological defects in tem-vivo role for NSF in neurotransmitter release. J. Neurosci. 18:10241–10249. perature-sensitive paralytic mutants of Drosophila melanogaster.

Proc. Natl. Acad. Sci. USA 73: 3253–3257.Koenig, J. H., and K. Ikeda, 1989 Disappearance and reformationof synaptic vesicle membrane upon transmitter release observed Stowers, R. S., and T. L. Schwarz, 1999 A genetic method for

generating Drosophila eyes composed exclusively of mitoticunder reversible blockage of membrane retrieval. J. Neurosci. 9:3844–3860. clones of a single genotype. Genetics 152: 1631–1639.

Stowers, R. S., L. J. Megeath, J. Gorska-Andrzejak, I. A. Meinertz-Leventis, P. A., B. M. Chow, B. A. Stewart, B. Iyengar, A. R.


hagen and T. L. Schwarz, 2002 Axonal transport of mitochon- Yun, B., R. Farkas, K. Lee and L. Rabinow, 1994 The Doa locusencodes a member of a new protein kinase family and is essentialdria to synapses depends on milton, a novel Drosophila protein.

Neuron 36: 1063–1077. for eye and embryonic development in Drosophila melanogaster.Suzuki, D. T., 1970 Temperature-sensitive mutations in Drosophila Genes Dev. 8: 1160–1173.

melanogaster. Science 170: 695–706. Zallen, J. A., S. A. Kirch and C. I. Bargmann, 1999 Genes requiredVaessin, H., E. Grell, E. Wolff, E. Bier, L. Y. Jan et al., 1991 pros- for axon pathfinding and extension in the C. elegans nerve ring.

pero is expressed in neuronal precursors and encodes a nuclear Development 126: 3679–3692.protein that is involved in the control of axonal outgrowth in Zelhof, A. C., H. Bao, R. W. Hardy, A. Razzaq, B. Zhang et al., 2001Drosophila. Cell 67: 941–953. Drosophila Amphiphysin is implicated in protein localization and

Wolff, J. R., W. L. Liu, H. Bottcher, I. Krizbai, F. Joo et al., 1997 membrane morphogenesis but not in synaptic vesicle endocytosis.Non-conventional role of lysosomal acid phosphatase in olfactory Development 128: 5005–5015.receptor axons: co-localization with growth-associated phospho-protein-43. Neuroscience 79: 887–891. Communicating editor: R. S. Hawley

Documents

A Genetic Screen for Synaptic Transmission Mutants Mapping to the