Developmental consequences of awdb3, a cell-autonomous lethal mutation of Drosophila induced by hybrid dysgenesis

DEVELOPMENTALBIOLOGY

129,159-168(1988)

Developmental Consequences Mutation of Drosophila CHARLES Department

R.

DEAROLF,~

of Biology,

of awdb3, a Cell-Autonomous Induced by Hybrid Dysgenesis EVELYN

The Johns

HERSPERGER,

Hopkins

Accepted

University,

May

AND Baltimore,

ALLEN Maryland

Lethal

SHEARN 21218

2, 1988

In order to recover mutations affecting imaginal discs in a way which would allow the relevant genes to be readily cloned, a hybrid dysgenic screen was performed for mutations causing late larval/early pupal lethality. This paper describes that mutagenesis procedure and the phenotypes caused by the mutations that were recovered. Of 81 late larval/pupal lethal mutations that were recovered, 20 cause imaginal disc defects. These 20 mutations define 12 different genes. This paper also includes a description of the developmental defects caused by a mutation in one of those 12 genes which we have named abnormal wing discs (awd); the following paper (C. Dearolf, N. Tripoulas, J. Biggs, and A. Shearn, 1988, Dev. Biol. 129, 169-178) describes the cloning of the awd gene and an analysis of its pattern of transcription. awdbS homozygotes develop at a normal rate until the end of the second larval instar, when their rate of development is reduced. After an extended third larval instar, they form puparia and die. Mutant wing discs have an abnormal morphology and extensive cell death. These abnormal wing discs, and also the leg and eye-antenna discs which appear to be morphologically normal, differentiate poorly or not at all when injected into metamorphosing larvae. Analysis of genetic mosaics indicates that the awd” mutation is expressed in a cell-autonomous manner in wing, leg, and eye-antenna discs. The larval brain and proventriculus in awd” homozygous third-instar larvae appear to be vacuolated due to the accumulation of lipid droplets. Mutant ovaries are unable to develop when injected into wild-type o I988 Academic PEW, IX larvae, although mutant germ cells are capable of producing normal eggs. INTRODUCTION

The imaginal discs of Drosophila melanogaster are structures in the larvae which form the epidermis of the adult head, thorax, and genitalia. We are using imaginal discs to examine how the products of specific genes are integrated into developmental processes. Genes necessary for the normal development of imaginal discs can be identified by mutations which cause one or more pairs of discs to develop abnormally. Two basic approaches have been used to screen for such mutations. One approach has been to isolate lethal mutations, determine which of them causes lethality after the third larval instar, and examine those for imaginal disc defects (Shearn et al., 1971; Stewart et al., 1972; Kiss et al., 1976). The other approach has been to isolate lethal mutations and examine them for autonomous imaginal cell defects in mosaics produced by somatic recombination (Rip011 and Garcia-Bellido, 1973; Russell, 1974; Arking, 1975; Simpson and Schneiderman, 1975). Both approaches have yielded mutations in many genes. A wide variety of techniques has been applied to study the developmental consequences of such mutations. In a few cases, the results of those studies have provided clues as to the biochemical function of the i Present address: Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA 91125. 159

affected gene products. For example, Cheney et al. (1984) showed that one function of the protein coded by the gene identified by 1(3)c21 (Pentz and Shearn, 1979) is to modify the structure of three abundant proteins. However, in general the developmental consequences of lethal mutations do not provide useful clues as to the biochemical function of the products of the genes which those mutations serve to identify. In order to take advantage of the techniques of molecular cloning for discovering the products of genes identified by lethal mutations, we have undertaken a screen for hybrid-dysgenesis-induced mutations which cause imaginal disc defects. Hybrid dysgenesis (Kidwell et ah, 1977) is a syndrome which occurs in the germ cells of progeny derived from mating males of certain wild-type strains (called P strains) to females of laboratory strains (called M strains). One symptom of this syndrome is the induction of mutations caused by the insertion of a transposable P element (Rubin et aZ., 1982). Genes tagged by such an element can be cloned by a straight-forward set of procedures. This paper describes the mutagenesis screen we used and summarizes the properties of the mutations we have recovered. Our results indicate that the phenotypes of these dysgenic mutants are similar to those of mutants derived from chemical mutagenesis. We have chosen one of the genes identified by these dysgenic mutations to study in detail. The objective of the work 0012-1606188 Copyright All rights

$3.00

0 1988 by Academic Press, Inc. of reproduction in any form reserved.

160

DEVELOPMENTAL

BIOLOGY

described in this paper was to infer as much as possible about the temporal and tissue-specific requirements for the product of this gene, called abnormal wing discs (awd), based on the developmental consequences of a dysgenic allele. The following paper describes the cloning of the awd gene and an analysis of the normal temporal and tissue-specific pattern of transcription of the gene. MATERIALS

AND

All cultures were maintained on a medium of cornmeal, yeast, molasses, and agar, at 20°C unless otherwise noted. The genetic markers and balancers used are described in Lindsley and Grell (1968). of Mutations

Mutations were induced by P-element hybrid dysgenesis and were recovered in a screen for third-chromosome, recessive, lethal, and visible mutations as shown in Fig. 1. Five different P strains were used to

II

+ TM3,

e Ser

X8

129,1988

generate mutations because P elements from different strains are thought to insert with different specificities (W. Engel, personal communication). Care was taken to maintain the potentially mutagenized chromosomes in flies having a P cytotype in order to minimize the likelihood of reversions caused by the excission of inserted elements. The markers used included multiple wing hairs (mwh, 3-O.O),Glued (Gl, 3-41.4), Stubble (Sb, 3-58.2), red Malpighian tubules (red, 3-53.6), ebony (e, 3-70.7), Serrate (Ser, 3-92.5), and lethal (3rd chromosome) Dominant Temperature Sensitive-2 (DTS-2,3-33.4, isolated by

METHODS

stocks

Generation

VOLUME

mwh +red

e

Holden and Suzuki, 1973). The balancer chromosomes used were TM3, e Ser and TM3, Sb e Ser. The P stocks employed were 7r2, Inbred Cage 3, 78.25, 78.100 and 8.31.15. P cytotype marker strains were generated by backcrossing flies with the desired genotype three to four times into known P strains. Lethal mutations on the third chromosome were kept as balanced stocks over TM3. For each of these stocks the stage of lethality of mutant homozygotes was determined. The imaginal disc morphology of homozygous larvae was examined for each of the mutations which caused lethality after the third larval instar. Deletions of the awd locus were isolated as y-ray-induced revertants of the dominant mutation, Killer of prune (K-pn, 3-102.9). The deletion used in this study, awdKR7, removes the entire awd gene (Biggs et al., submitted for publication). Viability and Life Span of awd” Larvae

mwh

‘” 9 TM3, V

mwh

red

e *

Sb e Ser

red

e *

TM3, Sb e Ser

Stock

mwh x

&

3

red

mwh

red

e *

mwh

red

e *

test

e *

TM3, Sb e Ser

for

lethality

FIG. 1. The mutagenesis screen utilized for obtaining late larval/ early pupal recessive lethal mutations. Cross I: Females of M-cytotype carrying three third-chromosome, recessive markers were mated to wild-type males of P-cytotype. Cross II: Dysgenic male progeny which were heterozygous for the three markers were mated to P-cytotype females carrying a third-chromosome halancer. Cross III: Individual female progeny heterozygous for a marked chromosome carrying a putative mutation (indicated by asterisk) and a balancer were mated to P-cytotype males heterozygous for a dominant mutation and a balancer containing the marker Stubble (Sb). Cross IV: Progeny heterozygous for the marked chromosome and the halancer carrying Sb were mated to each other. Cross V: If no progeny homozygous for the marked chromosome survived, then the heterozygous progeny were maintained as a stock.

Eggs were collected in uncrowded vials, counted, and allowed to develop for varying durations. Beginning at the second instar, which is the earliest stage at which the marker red is expressed, all of the larvae from several vials were counted and staged by the morphology of their anterior spiracles and/or of their mouth hooks (Bodenstein, 1950). In this and all future experiments, the awd stock used was mwh red e awdbS/TM3, Sb e Ser unless otherwise indicated. Histology

Mutant and control (mwh red e) larvae were treated in one of two ways: (1) Larvae were fixed in CarnoyLebrun fixative, cleared with dioxane, embedded in Paraplast, cut into 5-pm serial sections, and stained by the Feulgen/fast green method, which stains DNA redviolet and other tissue elements green. (2) Larvae were fixed in 10% formalin, frozen in OCT medium, cut at -18°C into 12-pm sections, and stained with Sudan III and Harris hematoxylin, which causes lipids to appear orange-red and other tissue elements to appear blue. These methods are described in Humason (1972).

DEAROLF, HERSPERGER, AND SHEARN

awdb” Imaginal

Disc Phenotype

Developmental

Consequences

of

awdhP

161

larvae (96-108 hr after egg laying) received a total of 2250 R. Adults of the genotype mwh red e awdti/mwh Ki Sb M(3)w lz4 or mwh red e/mwh Ki Sb M(3)w’” were scored for the presence and morphology of clones of Ki+Sb+ bristles. Clones homozygous for Ki Sb M(3)wlz4 were not recovered because the Minute mutation is cell lethal when homozygous (Stern and Tokunaga, 1971).

The eye-antenna, wing, and leg discs were dissected from third-instar larvae and their morphology was examined. To assay for cell death, discs were stained for 5 min in a solution of 0.5% trypan blue in Drosophila Ringer’s solution. Trypan blue is a dye which is excluded from intact cells. Under these conditions, no Ovary Transplants staining was found in wild-type discs unless injured during the dissection. To determine the potential of the Ovary transplants were performed by the method of discs to differentiate adult cuticular structures, mutant Clancy and Beadle (1937). Mutant, mwh red e awdbs, or third-instar discs were injected into wild-type larval control, mwh, ovaries were dissected from third-instar hosts using the method described by Ursprung (1967). larvae and injected into female larvae homozygous for Resulting implants were recovered from the adult hosts f~(3)L8~“. The use of these hosts for ovary injections was after metamorphosis, mounted in Berlese’s medium described by Shearn et at. (1978a). Surviving hosts were (Humason, 1972), and examined with a light micromated at 27°C to males of the genotype mwh red e (for scope. mutants) or mwh (for controls), and the progeny were examined. Females homozygous for fs(3)L8t* do not proDisc Growth Studies duce viable eggs at this restrictive temperature, so any progeny obtained were derived from the injected In larvae. Normal wing discs were dissected from ovaries. Surviving females which did not produce pro5-day-old Canton-S larvae (early third instar) and were geny were dissected to examine for the presence and implanted into either awdbS or Canton-S larvae of the appearance of donor ovaries. same age. The discs were cultured in control hosts for 2 days or-in mutant hosts for 7 days, and then recovered. Pole Cell Transplants The extra period of culture in mutants was thought to be necessary to compensate for the extended third inPole cell transplants were done as described by star of mutant larvae. The discs were measured using Lawrence et al. (1983). Pole cells were removed from an ocular micrometer before implantation and after re- individual embryos resulting from the cross of mwh red covery, and the calculated area was used as an estimate e awdbs/TM3, Sb e Ser parents, and injected into emof disc size (Shearn et ak, 197813). Some of the recovered bryos derived from the cross of wild-type females to discs were implanted into metamorphosing larvae, as males hemizygous for Fs(l)K1237. Fs(l)K1237 is a domidescribed above, to assay their capacity to differentiate nant, female-sterile mutation causing females to be adult structures. agametic (Perrimon and Gans, 1983); therefore any In adults. Wing discs were dissected from either 5- to eggs laid by recovered female hosts had to be derived 6-day-old awdbg larvae or 5-day-old Canton-S larvae. from the transplanted pole cells. The female hosts that Their sizes were estimated as described above and they were recovered were individually mated to either mwh were injected into Canton-S adult virgin females, which red e or mwh red e awd&/TMl, Me’ males and their were subsequently mated. The discs were cultured for progeny were examined. The donor pole cells were hetperiods of time ranging from 1 to 4 weeks. Upon re- erozygous like their parents (mwh red e awd&/TM3, Sb e moval from their hosts, the sizes of the discs were again Ser), homozygous for mwh red e awd@, or homozygous estimated. Some were stained with 0.5% trypan blue to for the balancer TM3, Sb e Ser. In the case of hosts that detect the presence of dead cells and some were injected were mated to mwh red e males, the presence of both into metamorphosing larvae. red and Sb e Ser progeny indicated that heterozygous pole cells had been transplanted, while the presence of only red progeny indicated that homozygous mutant Genetic Mosaics pole cells had been transplanted. None of the fertile The progeny of mwh red e awdbS/TM3, Sb e Ser fe- female hosts that were recovered were derived from embryos transplanted with pole cells homozygous for males or mwh red e/TM& Sb e Ser females and mwh Ki Sb M(A’)w~‘~/TM~ males (Ki = Kinked, 3.47.6; M(3)wlzA the balancer TM3, Sb e Ser. That is apparently because TM3 causes lethality of pole cell derivatives when ho= Minute(3)w 124, 3-79 .7) were grown at 25°C and irradiated with y-rays from a cesium source at a dose rate mozygous. In the case of hosts that were mated to mwh Me’ males, the presence of Sb e Ser of 276 R/min. Early second-instar larvae (48-52 hr after red e awd@/TMl, and Me’ adult progeny indicated that the host had been egg laying) received a total of 1750 R, while third-instar

162

DEVELOPMENTAL

BIOLOGY

injected with heterozygous pole cells, while the presence of Me’ adult progeny and the absence of Sb e Ser adult progeny indicated that the host had been injected with mwh red e awdbS homozygous pole cells. RESULTS

Generation

of Hybrid-Dysgenic

Mutants

Of 4680 third chromosomes which were screened for recessive lethality or non-lethal defects, 482 chromosomes with mutations were recovered (Table 1). Of these, 92 caused non-lethal defects: 85 caused delayed development and ‘7 caused morphological defects. The remaining 390 mutations caused lethality. This frequency of lethal mutations, 8%, is similar to that reported for the X chromosome (Simmons and Lim, 1980). Of the 390 lethals, 81 died after the third larval instar but before metamorphosis. Homozygous, third-instar larvae from each of these late larval/prepupal lethals were dissected to examine their imaginal disc morphology. Twenty of the eighty-one late larval/prepupal lethals caused obvious imaginal disc defects. A variety of defects was observed, ranging from all of the discs missing to only a single pair of discs appearing to be defective. This variety of defects is similar to that found in chemically induced mutants (Shearn et al., 1971; Stewart et ak, 1972; Kiss et al., 1976). The awdbg Mutation

We considered it necessary to concentrate on a single gene so that both the developmental and molecular consequences of mutations in that gene could be investigated in detail. The abnormal wing discs or awd gene was chosen for further study initially because only one pair of discs in awd b3 larvae appeared to be defective. This suggested that its gene product might have a tissue-specific function, in contrast to the products of most other genes which have been found to be essential for imaginal disc development (Shearn, 1978).

CLASSIFICATION Class

OF MUTATIONS Abnormality/stage

TABLE 1 OBTAINED

of lethality

Nonmutant

None

Nonlethal

Slow development Morphological abnormalities

Lethal

Embryonic lethal Early larval lethal Late larval/early pupal Late pupal lethal

Total

BY HYBRID

DYSGENESIS

Number

of stocks 4198 85 7

lethal

165 42 81 102 4680

VOLUME

129.1988

Viability and developmental rate. The lethal awdb3 mutation is recessive, maps near the end of the right arm of the third chromosome, and was recovered on a chromosome which did not contain any other lethal mutations (Dearolf, 1986). Larvae heterozygous for awdb3 have the same viability and developmental rate as wild-type controls. The frequency of awdbg homozygous larvae in cultures derived from heterozygous parents is 80% of that expected based on the number of heterozygous siblings; these homozygous larvae develop at a normal rate until the middle of the third instar, but then remain at this stage for several days or weeks before dying as prepupae. Imaginal disc morphology. The imaginal discs dissected from early, third-instar awdbS homozygotes appear morphologically normal. By mid third instar, however, most wing discs have abnormal morphologies. These wing disc morphologies range from slightly abnormal to grossly misshapen, as shown in Figs. 2b-d. Even within a single larva the two wing discs can have different morphologies. Wing discs from early thirdinstar larvae showed no signs of cell death; by the mid third instar, however, the wing discs showed extensive staining. In g-day-old larvae, 30/33 wing discs stained with trypan blue. Two patterns of staining were seen: in one pattern, a single large region of staining was observed in the disc, usually surrounding the region of the presumptive wing blade (Fig. 2e); in the other pattern, lighter staining regions were scattered throughout the disc (Fig. 2f). The other imaginal discs have a normal appearance at the mid third-instar stage and even during the extended third instar. None of the other discs in awdbY mid-third-instar larvae showed abnormal patterns of cell death. Only during the extended third instar, were scattered regions of light staining found in some of the other discs. In 13- to 15-day-old larvae, leg discs (12/17) and eye-antenna discs (6/6) stained lightly. No trypan blue staining was found in awdbS brains. Histology. In order to examine the cells of imaginal discs as well as those of other larval tissues at a higher level of resolution, mutant, third-instar larvae were examined in histological serial sections (Fig. 3). While mutant imaginal disc cells appear normal, the ventral ganglion, brain lobes, and to a lesser extent the anterior part of the proventriculus have abnormal vacuolations. These vacuoles resemble the lipid droplets normally seen in both control and mutant fat bodies. To test whether the vacuoles in the mutant brain and proventriculus contained lipid, sections of frozen larvae were stained with Sudan III and hematoxylin. In control larvae there are some small lipid droplets in the central part of the brain lobes and ventral ganglion, while in mutant larvae there are extensive, larger lipid droplets

DEAROLF,

HERSPERGER,

AND SHEARN

Developmental

throughout olf, 1986).

163

Consequences of awdbS

the brain lobes and ventral ganglion

(Dear-

Growth of mutant wing discs in wild-type adults. After wild-type 1 week of in vivo culture, early third-instar

FIG. 2. Morphology of normal and mutant wing discs. (a) Canton-S third-instar wing disc. (b-d) awdbS third-instar wing discs. (e-f) Mutant wing discs stained with trypan blue. (e) This disc has one large area of intense stain. (f) This disc has two areas of diffuse light stain. Magnification is 140X.

FIG. 3. Histological sections of awdbS larvae. (a) Feulgen/fast green staining of mutant mid-third-instar larva. (B) Brain, is X76.

discs grew to a mean size comparable to wing discs dissected from wild-type larvae at the end of the third larval instar (Fig. 4.). Longer durations of culture did not lead to significantly more growth. By contrast, early third-instar mutant discs even after 4 weeks of culture were significantly smaller than wild-type discs (P < 0.05 according to t test). Although able to grow, the mutant discs increased only gradually in size throughout the culture period. The cultured discs were examined for cell death by trypan blue staining. Mutant cultured discs stained as extensively as those removed from g-day-old third-instar awdb3 larvae (Figs. 2e,f), whereas control cultured discs had only a few areas of diffuse stain. This result suggests that the reduced net rate of mutant disc growth could be explained by an increased extent of cell death during in viva culture. Ll$Ferentiation of mutant discs. To test the ability of discs from homozygous awd bs discs to differentiate, third-instar mutant larvae were implanted into metamorphosing wild-type larvae. The resulting implants were examined for the presence of adult cuticular structures (Table 2). Most of the control discs (mwh red e) gave rise to all of the adult cuticular structures which are characteristic for each kind of disc whereas mutant poorly or not at discs (mwh red e awdbg) differentiated all. For example, only 1 of 20 mutant wing disc implants contained all of the characteristic wing structures, whereas 19 of 21 normal wing disc implants contained all of the characteristic wing structures. Capacity of awdbSlarvae to support the development of normal wing discs. Imaginal discs rely on larval tissues

for nutrients

green staining (P) proventriculus,

and hormones

which are necessary for

of mwh red e normal mid-third-instar (E) eye disc, (A) antenna1 disc,

larva. (b) Feulgen/fast (F) fat body. Magnification

164

DEVELOPMENTAL BIOLOGY

2

1 DURATION

3

OF CULTURE

4 (weeks)

FIG. 4. Growth of mutant B and Canton-S m wing discs in Canton-S adult hosts. Relative growth is defined as the increase in size of each disc following in &no culture divided by its size prior to culture. The disc sizes were estimated as described under Materials and Methods. Error bars indicate f one standard deviation of the mean.

their growth and development. It was possible that the imaginal disc abnormalities that had been observed were a secondary consequence of a primary defect in one of those larval tissues. To test this possibility, Cunton-S second-instar wing discs were implanted into either awdbS (n = 33) or Canton-S (n = 21) larvae of a similar age, and allowed to develop. It was found that normal wing discs can continue to grow in awdbg hosts (Dearolf, 1986) and those cultured in awdbg hosts (n = 8) were able to differentiate as well as those cultured in nonmutant hosts (n = 9). Genetic mosaics. Genetic mosaics were used to test whether the disc abnormalities caused by the awdb3 mutation are also cell autonomous, i.e., whether they result directly from the lack of the awd’ gene product in imaginal disc cells. The head, antennae, legs, dorsal

VOLUME

129,1988

prothorax, and mesothorax of each relevant fly were examined for the presence of KPSb’ clones. For larvae irradiated during the second instar, no homozygous awdbg clones were recovered on the head, antennae, legs, or the dorsal prothorax (Table 3). On the dorsal mesothorax which is derived from part of the wing disc a few clones were recovered but the frequency at which they were recovered was much lower than that of controls. Moreover, the morphology of bristles in the awdbg dorsal mesothoracic clones was abnormal in most cases; many of these bristles were thin, short, or bent, and in two cases bristles were duplicated. For larvae irradiated during the third instar, homozygous clones were recovered on the head, antennae, legs, dorsal prothorax, and mesothorax, but the frequency at which they were recovered was significantly lower than that of controls (Table 3). Also, the average size of the homozygous awdb3 clones was smaller than that of control clones. There was no indication of a preferred location of those clones that were recovered. Development of awd ovaries. The developmental capacity of awdbg mutant ovaries was tested by transplanting ovaries from mutant larvae into female host larvae. None of the 62 hosts which received awdbg donor ovaries gave rise to progeny. By itself, this result is of marginal significance since only 8% (4/49) of the hosts which received wild-type control ovaries gave rise to progeny. However, examination of hosts which failed to produce progeny revealed that none of 18 females which received mutant donor ovaries had an extra ovary whereas, 16 of 28 females which received wild-type donor ovaries did have an extra ovary. Those extra ovaries were well developed but did not give rise to progeny because they had not become connected to the hosts’ reproductive tract. This result suggests that

TABLE 2 DIFFERENTIATION

OF MUTANT

IMAGINAL

DISCS

Number of implants with”

Disc tested

Genotype

Total number of implants

All characteristic structures

0 8

67

8

0

3 10 0 10 1

21 0 6 0 7

0 0 5 0 3

mwh mwh

red e red e awd”

21 20

19 1

Leg

mwh mwh

red e red e awd”

75

24 10

mwh mwh

red e red e awd b9

Antenna

mwh mwh

red e red e awd”

11 10 11

No characteristic structures

2 11

Wing

Eye

Some characteristic structures

a For each normal imaginal disc there is an inventory of characteristic structures such as cuticular patterns, bristles, hairs and sense organs which are recognizable in differentiated implants. For each of the mutant discs, the recovered implants were examined for those characteristic structures.

DEAROLF,

HERSPERGER,

AND

Developmental

SHEARN

TABLE SOMATIC

RECOMBINATION

Number of Ki+Sb+ clones” Stage of irradiation Early second instar

Early third instar

Tissue examined

Control

Mutant

Cmequences

165

of awd”

3 IN MUTANT

FLIES

Average size of Ki+Sb+ clones b

Frequency of Ki+Sb+ clones Control

Mutant ii

Head Antenna Leg D. mesothorax D. prothorax

25 9 35 16 8

0 0 0 3 0

0.174 0.063 0.243 0.111 0.056

Od 0.022d Od

Head Antenna Leg D. mesothorax D. prothorax

33 16 32 20 5

16 7 7 15 1

0.559 0.272 0.542 0.338 0.085

0.225d 0.098” 0.098d 0.211” 0.014”

Control 11.8 6.6 23.9 30.9 4.0

Mutant

f + + + +

11.8 2.8 18.3 31.3 1.0

30.3 f 0.75 -

2.8 f 3.0 f 9.2 f 5.8 + 1.2 f

1.9 0.8 8.8 7.9 0.6

2.1 f 1.5 1.9 + 0.P 2.7 * l.ld 2.8 + 2.5” 1.0 f 0.5

“The numbers of flies examined were second instar control, 144; second instar mutant, 131; third instar mutant, 59; third instar mutant, 71. b Average clone size is the number of non-Ki Sb bristles in the clone c Significantly less than control (P < 0.05). d Significantly less than control (P < 0.01).

ovaries are not capable of normal development. The ovaries are composed of cells of somatic and of germ line origin. To test whether the awd& mutation causes autonomous defects in the germ line, pole cell transplants were performed. Out of 63 hosts mated to nonmutant males, 8 injected with pole cells from awdbS heterozygotes and 5 injected with pole cells from awdbg homozygotes were recovered. The progeny from the homozygous pole cells were viable and appeared to be normal. This result indicates either that the awdbsmutation does not cause an autonomous germ cell defect or that the defect could be rescued by awd+ sperm. To distinguish between these possibilities, another set of pole cell transplants was performed but in this set the hosts that were recovered were mated to males that were heterozygous for awdbg and TMl. The hosts which received homozygous donor pole cells could be distinguished because they gave rise to progeny with the balancer chromosome from the father, i.e., TMl, but did not give rise to progeny with the balancer chromosome that was present in heterozygous donor cells, i.e., TM3. Two such hosts with fertile homozygous donor cells were recovered. The phenotype of awdb” homozygous larvae derived from such homozygous female germ cells was identical to the phenotype of awdbg homozygous larvae derived from heterozygous female germ cells. This indicates that the awd” mutation does not cause an autonomous germ cell defect. Phenotype of awdbg hemixygotes. As a genetic test of whether the phenotype of awdbShomozygotes is caused by the absence (as opposed to the reduction) of awd+ function, the phenotype of awdb8 homozygous larvae awdbg mutant

was compared to that of awdbShemizygous larvae, i.e., larvae transheterozygous for awdb” and awdKRr, a deletion which removes the awd gene (Biggs et al., submitted for publication). It was found that the phenotype of mutant hemizygotes was indistinguishable from that of mutant homozygotes. Hemizygous larvae have an extended third larval instar; their wing discs are morphologically abnormal and they are unable to differentiate a complete inventory of characteristic imaginal structures. DISCUSSION

Chemically induced late larval/early pupal lethal mutations of Drosophila have identified many genes which are essential for normal imaginal disc development (Shearn et al., 1971; Stewart et al., 1972; Shearn and Garen, 1974; Kiss et aZ., 1976). The mutations recovered in this study cause imaginal disc abnormalities similar to those induced by chemical mutagenesis. The main advantage of mutations induced by hybrid dysgenesis is that these often result from the insertion of a transposed P element (Rubin et aL, 1982) which makes it convenient to clone the gene identified by such a mutation. Imaginal Disc Defects Caused by the awdbSMutation Are Cell Autonomous

The awd gene was initially chosen for further study because the awdbs mutation appeared to affect only wing discs. This would have been unusual, since most mutations that affect imaginal disc development, affect more than a single pair of discs and also affect larval

166

DEVELOPMENTALBIOLOGY V0~~~~129,1988

development (reviewed by Shearn, 1978). However, we subsequently found that other imaginal discs and some larval tissues are also affected by awdb? Mutant eyeantenna and leg discs, although morphologically normal, are not capable of differentiating like normal imaginal discs when cultured in metamorphosing larval hosts. Histological studies revealed that the brain ventral ganglion, and proventriculus have abnormal accumulations of lipid droplets. The observation of larval defects in addition to imaginal disc defects raised the possibility that the observed imaginal disc defects were a secondary consequence of a defective larval environment. This possibility was addressed in two ways: by culturing normal wing discs in mutant larvae and by producing clones of homozygous mutant cells in heterozygous, nonmutant, hosts. Normal wing discs when cultured in mutant larvae continued to grow as they did when cultured in normal larvae. Moreover, after culture in mutant larval hosts, those normal wing discs were capable of differentiating a complete inventory of wing structures. This demonstrated that mutant larvae could provide the cultured nonmutant discs with whatever substances wing discs normally require and did not provide any substance which interfered with normal development. These results suggested that the imaginal disc defects in the mutant resulted from the autonomous expression of the mutation in imaginal disc cells. This hypothesis was confirmed by analyzing genetic mosaics. Clones of homozygous cells on the head, legs, and thorax produced by somatic recombination during the second larval instar were not recovered at the frequency expected based on the results of control experiments (Table 3). This failure of homozygous cells to contribute to the adult cuticle is the operational definition of a cell-autonomous lethal mutation (Demerec, 1936). Many other autonomous cell-lethal mutations have been described (Rip011 and Garcia-Bellido, 1973; Russell, 1974; Arking, 1975; Simpson and Schneiderman, 1975; Stewart et aZ., 1972). For some of these mutations, evidence of cell death in imaginal discs during the larval period has been presented (Murphy, 1974; Clark and Russell, 1977). In the case of awdbg wing discs as well, there is evidence of cell death beginning at the middle of the third instar, which is when awdb” wing disc morphology begins to appear abnormal. So, cell death may be responsible for the abnormal wing disc morphology. If so, the requirement for awd+ must be different in the eye-antenna and leg discs than it is in wing discs. The morphology of mutant eye-antenna and leg discs is normal, even though they, like mutant wing discs, cannot give rise to characteristic imaginal structures when cultured in metamorphosing wild-type larvae.

awdM is a Null Activity

Allele

Since one goal of this project was to infer the temporal and tissue-specific requirements for the awd+ gene product from the developmental consequences of the awdbg mutation, it was important to try to discover whether awdbS was a null activity allele. This was done by comparing the phenotype of awdbg homozygotes to that of awdbg hemizygotes. The rationale of this test is that a mutation which is a null activity allele should cause the same phenotype regardless of whether one or two doses of the mutant allele is present. The observation that the phenotype of awdb” homozygotes was indistinguishable from that of hemizygotes, implies that awdbg is in fact a null allele. This was surprising to us since we originally interpreted the variability of the homozygous phenotype, as illustrated for example in Figs. 3b-d, as evidence that awdbg was a leaky or hypomorphic allele. Others, working on a different organism, have also reported phenotypic variability caused by apparent null mutations (Ferguson et ah, 1987). The absence of any normal transcripts of the awd gene in mutant larvae provides a molecular explanation of why awdbY is a null activity allele (Dearolf et ah, 1988). Temporal Specificity Product

of the Requirement

for awd+

Early third-instar larvae homozygous for awdbs appear phenotypically normal. Up until this stage they develop at the same rate and with the same viability as their heterozygous sibs. The lack of early lethality could indicate that the product of this gene is not required early in development but it could also indicate that the early requirement is met by maternal expression of the gene. To distinguish between these possibilities we performed pole cell transplants to produce progeny derived from a homozygous female germline. The fact that the phenotype of awdbS homozygous larvae derived from these females was identical to that of awdbg homozygous larvae derived from heterozygous awdbg females implies that the maternal genome does not need to provide embryos with awd+ gene product in order for those embryos to develop to the end of the second instar. Therefore, the lack of early lethality of awdbg homozygous larvae does in fact indicate that the product of this gene is not required early in development. The results of analyzing genetic mosaics also support that conclusion. The fact that very few homozygous clones were recovered following irradiation of heterozygous larvae during the second instar (Table 3) implies either that awd’ had not yet begun to be expressed in imaginal cells or that the product had not accumulated to a sufficient level to allow development to continue in

DEAROLF,

HERSPERGER,

AND

SHEARN

the absence of a functional awd’ gene. The few clones that were recovered were all derived from wing disc cells suggesting perhaps that awd+ expression begins earlier in wing discs than it does in other discs. The fact that relatively more homozygous clones were recovered following irradiation of heterozygous larvae during the third larval instar (Table 3) implies that by this stage many imaginal cells have accumulated a sufficient level of awd+ product that they can continue to develop in the absence of a functional awd+ gene. This evidence suggests that expression of the awd’ gene is no longer essential after the third larval instar. Tissue SpeciJicity of the Requirement for awd’ Product Imaginal discs. The abnormal morphology of homozygous awdbgwing discs, the inability of mutant wing discs to differentiate characteristic imaginal structures, and the cell-autonomous expression of the mutant phenotype indicate that expression of awd’ is essential for normal wing disc development. The inability of mutant eye-antenna and leg discs to differentiate characteristic imaginal structures and the cell-autonomous expression of the mutant phenotype in these discs indicate the awd+ expression is also essential for the development of these discs. However, the normal morphology of eye-antenna and leg discs suggests that in some way the function or the extent of awdt activity in those discs is not identical to its function or extent in wing discs. Larval organs. Histological studies of awdb3 larvae revealed numerous holes in the brain and ventral ganglion. These proved to be lipid-filled vesicles. This appears to indicate that awdt expression is also necessary in the larval nervous system. The product of the awd+ gene does not appear to be essential for the development of most other larval organs. This conclusion is based not only on the normal morphology of homozygous mutant larvae but more significantly on their ability to complete larval development including the formation of puparia. Ovary. A large number of zygotic, cell-autonomous lethal mutations identify genes which are also required for female fertility (Russell, 1974; Arking, 1975). In principle, those genes can be divided into three classes: those required for the somatic cells of the ovary, those required for the germ cells of the ovary, and those required for both the somatic and germ cells of the ovary. In order to determine in which, if anv, of these classes the awd gene belongs we performed ovary and pole cell transplant experiments. The ovary transplants have shown that the awd’ product is indeed required for the normal development of the ovary. The pole cell trans-

Developmental

Cmequences

of awdbY

16’7

plants have shown that this effect is not germ-line autonomous. These data indicate that the product of the awd’ gene is necessary for the development of the somatic cells of the ovary, but not for the germ cells. However, the data from these transplantation experiments do not address the possibility that the awd’ product which is made and required in the somatic cells of the ovaries is also transported into developing oocytes. Taken together all of these results indicate that there is no zygotic requirement for the awdf gene during embryonic or early larval development, and toward the end of the second larval instar or the beginning of the third larval instar the awd’ gene begins to be required for the development of the wing, eye-antenna, and leg imaginal discs, and parts of the brain and proventriculus, and the ovaries. The awdbgmutation was caused by the insertion of a P element. The presence of that element has allowed us to isolate the DNA sequence which corresponds to this mutant gene (Dearolf et ak, 1988). The cloned gene has allowed us to test the validity of the conclusions reported here about expression of the awd+ gene based on the developmental consequences of the awdb3mutation. This work was supported by grants to A.S. from the National Science Foundation and the National Institutes of Health. We thank Amanda Simcox and Nick Tripoulas for valuable discussions and William Engel for fly stocks. REFERENCES ARKING, R. (1975). Temperature-sensitive cell lethal mutants of Drosophila: Isolation and characterization. Genetics 80, 519-537. BODENSTEIN, D. (1950). The postembryonic development of Drosophila. In “Biology of Drosophila” (M. Demerec, Ed.), pp. 2’75-36’7. Wiley, New York. CHENEY, C. M., MILLER, K. G., LANG T. J., and SHEARN, A. (1984). Specific protein modifications are altered in a temperature-sensitive Drosophila developmental mutant. Proc Natl. Acad Sci. USA 81,6422-6426. CLANCY, C. W., and BEADLE, G. W. (1937). Ovary transplants in Dro sophila

melanogaster.

Biol.

Bull

72,47-56.

CLARK, W. C., and RUSSELL, M. A. (1977). The correlation of lysozomal activity and adult phenotype in a cell-lethal mutant of Dromphila. Dev.

Biol.

57,160-173.

DEAROLF, C. (1986). A phenotypic and molecular analysis of the L?ro sophila awd locus. Ph.D. dissertation, The Johns Hopkins University, Baltimore, MD. DEAROLF, C., TRIPOLJLAS,N., BIGGS, J., and SHEARN, A. (1988). Molecular consequences of awdm, a cell-autonomous lethal mutation of Drosophila induced by hybrid dysgenesis. Dev. Biol. 129,169-178. DEMEREC, M. (1936). Frequency of “cell-1ethals” among lethals obtained at random in the X-chromosome of Drosophila melanogaster. Proc.

Natl.

Acad.

Sci. USA 22,350-354.

FERGUSON, E. L., STERNBERG, P. W., and HORVITZ, H. R. (1987). A genetic pathway for the specification of the vulva1 cell lineages of Caenorhabditis eleaans. Nature (London) I 326.259-267. I HOLDEN, J., and SUZUKI, D. T. (1973). Temperature-sensitive muta-

168

DEVELOPMENTAL BIOLOGY

tions in Drosophila tal characteristics

melanogaster. XII. The genetic and developmenof dominant lethals on chromosome 3. Genetics

73,445-458.

HUMASON, G. L. (1972). “Animal Tissue Techniques,” 3rd ed. Freeman, San Francisco. KIDWELL, M. G., KIDWELL, J. F., and SVED, J. A. (1977). Hybrid-dysgenesis in Drosophila melanogastw: A syndrome of aberrant traits including mutation, sterility, and male recombination. Genetics 86, 813-833. KISS, I., BENCZE, G., FEKETE, E., FODOR, A., G~usz, J., MAROY, P., SZABAD, J., and SZIDONYA,J. (1976). Isolation and characterization of X-linked lethal mutants affecting differentiation of the imaginal discs in Drosophila melanogaster. Theor. AppL Genet. 48,217-226. LAWRENCE, P. A., JOHNSON, P., and STRUHL, G. (1983). Different requirements for homeotic genes in the soma and germ line of Dro sophila.

Cell 35, 27-34.

LINDSLEY, D., and GRELL, E. H. (1968). “Genetic Variations of Dro sophila melanogaster.” Carnegie Institute of Washington Publication 627. MURPHY, C. (1974). Cell death and autonomous gene action in lethals affecting imaginal discs in Drosophila melanogaster. Dew. BioL 39, 23-36.

PENTZ, E. S., and SHEARN, A. (1979). Analysis of the autonomy of imaginal disc defects in a small-disc mutant of Drosophila melane gaster.

Dev. BioL

70,149-170.

PERRIMON, N., and GANS, M. (1983). Clonal analysis of the tissue specificity of recessive female sterile mutations of Drosophila melanogaster using a dominant female sterile mutation Fs(l)K1237. Dew. BioL

100,365-373.

RIPOLL, P., and GARCIA-BELLIDO, A. (1973). Cell autonomous lethals in Drosophila melanogaster. Nature (London) 241,15-16. RUBIN, G. M., KIDWELL, M. G., and BINGHAM, P. M. (1982). The molecular basis of P-M hybrid dysgenesis: The nature of induced mutations. Cell 29, 987-994. RUSSELL, M. A. (1974). Pattern formation in the imaginal discs of a

VOLUME 129,1988 temperature-sensitive ter. Dev.

cell-lethal

mutant of Drosophila

melanogas-

BioL 40.24-39.

SHEARN, A. (1978). Mutational dissection of imaginal disc development. In “Genetics and Biology of Drosophila” M. Ashburner, and T. R. F. Wright, Eds.), pp. 443-510. Academic Press, London. SHEARN, A., and GAREN, A. (1974). Genetic control of imaginal disc development in Drosophila. Proc. Natl. Acad. Sci. USA 71, 1393-1397. SHEARN, A., HERSPERGER, G., and HERSPERGER, E. (1978a). Genetic analysis of two allelic temperature-sensitive mutants of Drosophila melanogaster both of which are zygotic and maternal-effect lethals. Genetics

89,341-353.

SHEARN, A., HERSPERGER, G., HERSPERGER, E., PENT& E. S., and DENKER, P. (1978b). Multiple allele approach to the study of genes in Drosophila melanogaster that are involved in imaginal disc development. Genetics 89,355-370. SHEARN, A., RICE, T., GAREN, A., and GEHRING, W. (1971). Imaginal disc abnormalities in lethal mutants of Drosophila. Proc. NatL Acad

Sci. USA 68.2594-2598.

SIMMONS, M. J., and LIM, J. K. (1980). Site-specificity of mutations arising in dysgenic hybrids of Drosophila melanogaster. Proc. NatL Acad

Sci. USA 77,6042-6046.

SIMPSON, P., and SCHNEIDERMAN, H. A. (1975). Isolation of temperature-sensitive mutations blocking clone development in Drosophila melanogaster and the effects of a temperature-sensitive cell lethal mutation on pattern formation in imaginal discs. Wilhelm Roux’s Arch.

Dev.

BioL

178,247-275.

STERN, C., and TOKUNAGA, C. (1971). On cell lethals in Drosophila. Proc.

NatL

Acad.

Sci. USA 68,329-331.

STEWART, M., C., MURPHY, C., and FRISTROM, J. (1972). The recovery and preliminary characterization of X-chromosome mutants affecting imaginal discs of Drosophila melanogaster. Dev. BioL 27, 71-83.

URSPRUNG,H. (1967). In vivo culture of Drosophila imaginal discs. In “Methods in Developmental Biology” (F. H. Wilt and N. K. Wessels, Eds.), pp. 485-492. Crowell, New York.