Copyright 0 1996 by the Genetics Society of America
Single Amino Acid Mutations in Drosophila Fascin Disrupt Actin Bundling Function in Vivo Kelly Cant and Lynn Cooley Department of Genetics, Yale University School of Medicine, New Haven, Connecticut 06510 Manuscript received October 19, 1995 Accepted for publication January 18, 1996 ABSTRACT Fascins bundle actin filaments into large, tightly packed hexagonal arrays that support diverse cellular processes including microvillar projections and filopodial extensions. In Drosophila, fascin is encoded by the singed locus. Severe singed mutants have gnarled bristles and are female sterile due to a defect in rapid cytoplasm transport duringoogenesis. In this paper, we report theresults of a large EMS mutagenesis screen to generate new singed alleles. A mutation that changes glycine 409 to glutamic acid results in partial inactivation of fascin in vivo; singedC409E mutants have kinked bristles and are fertile with a mild nurse cell cytoplasm transport defect. This mutation is in a small conserved domain near the C terminus of fascin. A mutation that changes serine 289 to asparagine almost completely inactivates fascin in vivo; singeds289Nmutants have gnarled bristles and are sterile due to a severe defect in nurse cell cytoplasm transport caused by the absence of nurse cell cytoplasmic actin bundles. A subsequent EMS mutagenesis screen for dominant suppressors of singeds289N sterility revealed an intragenic suppressor mutation that changes serine 251 to phenylalanine and restores much of fascin’s function. These two mutations, S289N and S251F, draw attention to a central domain in fascin.
A
CTIN bundles are necessary for the formation of specialized cellular processes, intercellular communication, and cell migration. Actin binding proteins guide the reorganization of the cytoskeleton that underlie these processes. Fascins are a newly emerging class of actin bundling proteins. Fascin was originally described as a 58-kD component of actin gels made from sea urchin egg extracts ( W E 1976). When purified, fascin was found to have actin bundling activity. Actin bundles containingfascin are hexagonally packed, have a characteristic cross-banding pattern of 12 nm periodicity, and contain a molar ratio of fascin to actin of 1:4.3 ( M E 1976;DEROSIER et al. 1977; BRYAN and W E 1978; SPUDICH and A M O S 1979;DEROSIERand CENSULLO 1981; CANT et al. 1994). Fascin was localized to actin filament bundles in both sea urchin egg microvilli and coelomocyte filopodial extensions (OTTO et al. 1979, 1980; OTTOand BRYAN 1980). Fascin has been implicated in the organization of actin filaments bundles in many organisms. When sea urchin fascin was cloned (BRYAN et al. 1993), the peptide sequence was found to have homologyto the Drosophila singed (sn)gene product (PATERSON and O’HARE 1991) and HeLa cell p55 (YAMASHIRO-MATSUMURA and MATSUM U M 1985, 1986). Drosophila singed protein bundles actin filaments with the same 12-nm periodicity described for sea urchin fascin and has been implicated in the formation of actin bundles in bristles and nurse
-
Corresponding author: Lynn Cooley, Department of Genetics, Yale University School of Medicine, 333 Cedar St., New Haven, CT 06510. E-mail:
[email protected] Genetics 143: 249-258 (May, 1996)
cells (CANT et al. 1994).Antibodies to p55 stain microspikes and stress fibers in both rodent and human cell lines (YAMASHIRO-MATSUMURA and MATSUMURA 1986; ADAMS 1995). Recently, mouse (EDWARDS et al. 1995), Xenopus (HOLTHIUS et al. 1994), and human (DUH et al. 1994; MOSIALOSet al. 1994) homologues offascin have been described. Using the algorithm, Pileup (UNIVERSITY OF WISCONSIN GENETICCOMPUTER GROUP 1995),fascin homologues are -35% identical and 60% similar, but amino acid conservation among fascin homologues does not reveal specific functional domain organization. Fascin shouldcontain two actin binding domains since the protein cross-links actin filaments as a monomer (BRYAN and KANE 1978; DEROSIER and CENSULLO 1981). Traditional biochemical approaches of mapping actin binding domains have had limited success. Truncated proteins expressed in Escherichia coli have either been insoluble or incorrectly folded (R. EDWARDS, personal communication). A human 27-kD C-terminal fragment (ONO et al. 1994) and a mouse 30-kD Gterminal fragment (EDWARDS and BRYAN1995) were generated using limited proteolysis and analyzed. The Gterminal half of human or mouse fascin appears to be able to bind, but not bundle, actin filaments and thereforecontains at least one actin binding domain. Genetic analysis can provide an in vivo complement to biochemical and structural approaches used to dissect the interactionbetween actin binding proteins and actin. In yeast, elegant genetic analysis involvinga combination of mutagenesis and suppressor screens defined the critical amino acids in both fimbrin and actin that
250
and
K. Cant
arerequiredforthe fimbrin-actin interaction. Mutations in fimbrin (SAC@can suppress mutations in actin (ACTI), and mutations in actin (ACTI) can suppress mutationsin fimbrin (SAC@ (ADAMS and BOTSTEIN 1989; ADAMS et al. 1989, 1991). This genetic evidence for physical interaction between fimbrin and actin was confirmed by actin bundling assays in vitro and protein colocalization in vivo (DRUBINet al. 1988). Molecular analysisof the mutation sites revealed the specific amino acids involved in this interaction and these findings were consistent with structural models (HONTSet al. 1994). Drosophila is an ideal organism for mutagenesis and genetic analysis. The Drosophila fascin homolog, singed (sn),has been well characterized. Strong sn mutants have gnarled bristles andare female sterile (MOHR 1922; BENDER1960). Bristles are formed as a single cellular extension that is supported by a cytoskeletal core composed of central microtubules surrounded by 8- 12 actin bundles that lie just beneath the membrane (OVERTON 1967; &PEL et al. 1993). In sn mutants, the actin bundles are very small and disorganized (OVERTON 1967; CANT et al. 1994), and they lack the hexagonal packing and 12-nm periodicity seen in wild-type bundles (TILNEY et al. 1995). Theseactin bundles probably lack thestructural integrity needed to promote straight bristle extension. Sterile s n females have a defect in the rapid phase of nurse cell cytoplasm transport (BENDER 1960). In nurse cells, two actin filament networks contribute to the rapid phase of nurse cell cytoplasm transport. The subcortical actin filament network supports myosin-based nurse cell contraction that pushes nurse cell cytoplasm intothe oocyte (GUTZEIT 1986; COOLEYet al. 1992; WHEATLEY et al. 1995). The cytoplasmic actin bundle network is thought to anchor the nurse cell nuclei in a central position away from ring canals (COOLEY et al. 1992; CANT et al. 1994; MAHAJAN-MIKLOS and COOLEY 1994). In egg chambers from sterile sn females, the subcortical actin network is apparently normal but the cytoplasmic actin bundle network fails to form; when rapid nurse cell cytoplasm transport initiates, untethered nuclei block the flow of cytoplasm through the ring canals resulting in the productionof small, infertile oocytes (CANT et al. 1994). More than 100 sn alleles have been described; however, few are candidates for mutations in the coding region. Many s n alleles are caused by transposable elements that typically insert into the promoter region, reducing sn transcription, Molecular analysisoffive spontaneous sn alleles identified mutations in the promoter or 5' untranslated regions of the gene (LINDSLEY and ZIMM 1992). EMS mutagenesis screens for female sterile mutations on the X chromosome yielded eight EMS s n alleles (GANSet al. 1975; MOHLER1977;KOMITOPOULOU et al. 1983). Unfortunately, only twoof these alleles are extant, M3 (MOHLER1977) and 1421 (KOMI-
L. Cooley et al. 1983). We conducted a large EMS mutagenesis screen to isolate additional EMS sn alleles. By analyzing EMS alleles of sn, we identified single amino acid changes that alter fascin function in vivo. We then used one of these new sn alleles in a second mutagenesis screen to identify a dominant suppressor of sn sterility.
TOPOULOU
MATERIALSAND METHODS Drosophila stocks: All fly stocks were maintained under standard culturing conditions. Canton S flies were used as the wild-type control. Males with an isogenic wl"' chromosome were maintained as a stock with C(l)yf/Y females. A FM7c snx2 stock was used to provide a nulls n allele (PATERSON and O'HARE 1991; CANT et al. 1994). f,kedj6" is a severe allele that was obtained from the Indiana University stock center et al. 1994). furrowed' was o b (HOOVERet al. 1993; PETERSEN tained from the Mid-America Drosophila Stock Center. Westernimmunoblotting: Drosophila ovaries were dissected from 4-5-day-old females that had been fed a yeast paste for 24 hr. Ovaries were ground in Laemmli sample buffer and boiled for 5 min. Extracts from one ovary per sample were separated by SDSPAGE (LAEMMLI 1970) and transferred to nitrocellulose membranes (TOWBIN et al. 1979). The nitrocellulose was immunoblotted with singed monoclonal antibody supernatant 7C as previously described (CANT rt al. 1994). 2D-Gel electrophoresis: For isoelectric focusing (IEF) , we used the method of O'FARREL.L (1975). IEFgels were prepared using 3.2% ampholytes, pH 6-8 (Bio-Rad), and 0.8% ampholytes, pH 3-10 (Bio-Rad). Gels were cast in 11.5 X 3cm diameter IEF tubes. Approximately 200 pg of total ovary protein was loaded onto the IEF gels and electrophoresed at 800 V for 18 hr and then 1800 V for 20 min. 10% SDS-PAGE gels were used forthe second-dimensionelectrophoresis (LAEMMLI 1970). Transfer to nitrocellulose was performed according to TOWBINand coworkers (1979). Nitrocellulose was immunoblotted with singed monoclonal antibody 7C as previously described (CANT et al. 1994). Northernanalysis: Three female flies for each sn allele were ground in 300 p1 of RNA extraction buffer (0.1 M NaCI, 0.5% SDS, 50 mM Tris pH 8, 10 mM EDTA, 50 &ml proteinase K) andincubatedfor 1 hr at 37". Samples were then extracted two times with phenol-chloroform (1:l) and once with chloroform. RNAwas ethanol precipitated and resuspended in 4.5 pl 1% diethylpyrocarbonate-treated (DEPC) water. RNA formaldehyde denaturation andgel electrophoresis were performed as described in XUEand COOI.EY(1993). RNA gels were blotted by capillary transfer onto Hybond-N (Amersham) and ultraviolet cross-linked to the membrane (Stratalinker, Stratagene). 92P-labeledprobes were prepared using random hexamer priming (FEINBERG and VOGEISTEIN 1983) of the 1.5-kb singed open reading frame and the open reading frame of the ribosomal protein rpAl as a loading control (QIAN et al. 1987).Filters were hybridized in Church's buffer at 65" with agitation (CHURCH and GILBERT 1984). Egg chamber staining procedure: Three- to 5day-old females were fed a water yeast paste for 24 hr. Ovaries were dissected and separated into individual egg chambers in icecold Drosophila EBR saline solution (130 mMNaC1, 5 mM KCl, 2 mM CaCI2, 10 mM HEPES, pH 6.9) as described (VERHEYEN and COOLEV 1994). Egg chambers were fixed in 6% formaldehyde and stained with rhodamine-phalloidin (Molecular Probes) or singed monoclonal antibody supernatant 7C as described (CANT et al. 1994). Confocal microscopy: Scanning laser confocal images were collected using the Bio-Rad MRCGOO system. All images were
Fascin
25 1
ofGenetic Analysis
collected using a Zeiss X25 lens with a numerical aperture of 0.8. Optical sections of 1-2 pm were combined using the CoMOS software package (Bio-Rad). Composites were assembled using Adobe Photoshop. Mutagenesis: Ethyl methane sulfonate (EMS) mutagenesis was carried out essentially as described (LEWISand BACHER 1968). Screen for new singed alleles (Figure 1): Zero- to four-day-old isogenic w1118maleswere starved for 12-16 hr and then fed 20 mM EMS in 1% sucrose for 16 hr. Flies were distributed into bottles with 10 mutagenized males and 40 virgin C(l)yf/ Y females. The flies were transferred to new bottles on days 2, 4, and 6. All males were removed before the final transfer on day 6 and the females were removed on day 8. All males progeny with bristle defects were selected and mated with C(l)yf/Y females. Independent lines were generated from males that could stably transmit the bristle defect to their progeny. Mutagenized X chromosomes were maintained in males by mating to C(l)yf/ Y females. Complementation testin with sn and forked ( f ) was performed using FM7c snx2and f' alleles, respectively. Those bristle mutantsthat complemented sn and f were placed in complementation groups. Female stocks were established by mating male bristle mutants with FM7a females. Females homozygous for bristle defects were tested for fertility by placing 20 females and 10 males in a vial and looking for larval progeny after 10 days. 6): Screen f o r dominant suppressors of sn stm'lity(Figure w I / / R snSZ#YN / Y males were fed 25 mM EMS in 1% sucrose for 16 hr. Flies were distributed into bottles either at 25 or 18". Each bottle contained 10 mutagenized males and 35 wll18sn,VZKY~V /FM7a balancer females. The flies were transferred to new bottles on day 2. On day 4, the males were removed and thefemales were placed in new bottles and then removed on day 7. Homozygous F1females were selected and tested for fertility. Any progeny obtained must have been derived from females carryin a dominant suppressor of the sn female sterility. F2w"18snsz 9N/w"'8sn~s289Nfemales were tested for fertility at 18' by mating to fresh w1118snS289N/Y males. An F2 w1118 sn""""/ Y males obtained were mated with stock ~ ~ ~ ~ s n " " ~ ~ / F M 7 n f e r n a lthe e s subsequent and F3w1118sns28yN/ ll18snS289N females were tested for fertility at 18". Molecularanalysis of sn alleles: All basic cloning techniques were performed as describedin SAMBROOK et al. (1989). Genomic DNAwas prepared from homozygous sn flies by standard techniques (protocol 48 in ASHBURNER 1989b). sn exons were amplified from genomic DNA using PCR primers designed to allow amplification of exon 2 and exons 3-6. PCR fragments were purified using PCR magic preps (Promega) or gel purified using Qiaex gel extraction kit (Qiagen). Overhangs (5') on PCR products were filled in using Klenow DNA polymerase (New England BioLabs). Blunt PCR products were subcloned into the EcoRV or the SmaI site of PCRscript (Statagene). Blue-white color selection was used to identify plasmids containing inserts. Insert-containing plasmids were sequenced using the USB sequenase kit. All mutations were confirmed by sequencing clones from independent PCR reactions.
B
B
RESULTS
E M S mutagenesis for new sn alleles: We performed an EMS mutagenesis screen to isolatenew sn alleles withmissense or nonsense mutations in the protein coding region. The mutagenesis screen was designed to identify genes affecting bristle morphology on the X chromosome (Figure 1).Flies were fed 20 mM EMS, a dose that is 5 mM less than is typically used, to reduce
Screen Results 68,500 Males
JI 0
48 Bristle mutants complementation testing
forked singed
J.
23f alleles 15 sn alleles 10 Others
FIGUREl.-EMS mutagenesis screen for new sin ed (sn) alleles was performed as diagrammed. Isogenic w"' males were mutagenized and matedto C(l)yf/ Y females. Male progeny with bristle defects were selected and independent lines were generated from those males that could stably transmit the bristle defect to their progeny. We tested new bristle mutants for complementation with sn and forked, two common bristle morphology mutants on the X chromosome.
male sterility and lethality. We screened -68,500 male progeny of mutagenized males and obtained 48 new bristle mutants on the X chromosome. Complementation analysis classified our 48 new bristle mutants as 15 sn alleles, 23 fwked alleles, and 10 mutants that complemented sn and fwked (Figure 1). One sn allele had an intermediate phenotype; the bristles weremoderately bent and thefemales werefertile. The remaining 14 s n alleles had severe phenotypes that resembled the null allele phenotype of snX2;they had gnarledbristles and the females weresterile. These alleles were not temperature sensitive. Complementation testing among the 10 non-sn, non-fwked mutants placed them into two groups, A and B. Complementation group A contained eight alleles that differed in severity and exhibited a variety ofphenotypes including short, gnarled bristles, rough eyes, and poor viability with increased pupal lethality. We mapped the mutation in complementation group A to 1.38.5 using meiotic recombination. furrowed maps to 1.36.85 (LINDSLEY and ZIMM 1992) and furrowed mutants have a similar spectrum of phenotypes. Complementation testing revealed that mutants in complementation group A were fuwowed alleles. Complementation group B contained two alleles that bothexhibited a subtle thin, bentbristle phenotype and one allele had reduced female fertility. This complementation group mapped roughly to 1.10, and we have not determined a previously identified locus affected in complementation group B. Analysis of EMS sn alleles: We were primarily interested in analyzing those s ? alleles ~ in which the observed phenotype could be explained by a defect in protein function rather than anabsence of protein. Therefore, we performed Northern and Western analysis on the 15 new EMS alleles and two previously described EMS sn alleles, M3 (MOHLER1977) and 1421 (KOMITOPOULOU et al. 1983). sn encodes three transcripts of 3 . 6 , 3.3, and 3.0-kb (PATERSON and O'HARE1991). These
K. Cant and L. Cooley
252
A.
and disorganized (Figure 3E). Notall of the nurse cell cytoplasm was transported into the oocyte and the eggs were -75% the sizeofwild-typeeggs (Figure 3H). While the actin defect in sn"'"" nurse cells was subtle, the actin defect in snS2XyNwas severe and appearedsimilar to the defect seen in null snx2 mutants (CANTet al. 1994); the subcortical actin was normal but the cytoplasmic actin bundles were nearly absent (Figure 3F) B. The rapid phase of cytoplasm transport in this allele was disrupted and the oocyte remained only half the size of a wild-type egg (Figure 31). The phenotypes of sn"'"'. and snsZXyN suggested that 69. 0 0 0 kelch the fascin expressed in these alleles was abnormal (Tam slngcd ble l ) . To identify the mutation, the sn open reading 47frame was amplified from genomic DNA isolated from FIGURE2.-Northern (A) and Western (B) analysis of EMS the parental line (isogenic z u r r r x ) ,~ n " ~ ~ )and ~ " ' sn , S28Y.v sn alleles. (A) Female total RNA was probed for sn and ?A1 and then sequenced. ~n,"~"~':' had a single nucleotide G (QIANet nl. 1987) as a loading control. Transcriptscorresponding to the three sn transcripts of 3.6, 3.3, and 3.0 kb + A transition in codon 409 that changed a GGA (glywere detected inwild type. sn transcriptswere detected in cine) to GAA (glutamic acid) (Figure 4). sns2X9,v had a only four EMS alleles, sn""'",; 5n1"', sn"', and sn~sz8y~v . The single nucleotide G A transition in codon 289 that smaller,ovary-specifictranscriptswere absent from ~ n " ~ . sn 1.121 and snnf3were generated in previous EMS mutagenesis changed AGC (serine) to AAC (asparagine) (Figure 4). The severe phenotypes of sn~s2Xy~v suggested that the screens (KOMITOPOUI.OU el nl. 1983 and MOHLER1977, respectively). Molecular size markers (2.2 and 4.4 kB) are indimutation in serine 289 dramatically altered fascin funccated on the left. (B) Ovary protein extractswereimmution. In snsZXyN mutants, the bristle phenotype was only noblotted withsn monoclonal antibody 7C for singed and slightly less gnarled than the severely gnarled phenowith kelch monoclonal antibody 1B as a loading control. Fastype of the null allele snxz and the egg chamber phenoand sn,sz8y." . Molecular weight cin was detected only in sn""f'yl'.' type mimicked that seen in null sns2 allele. Serine 289 markers (47 and 69 kD) are indicated on the left. was contained in a region with no homology to actin transcripts differ only in their polyadenylation site and binding domainsidentified in other proteins. Many isothe 3.3- and 3.0-kb transcripts are unique to ovaries. forms of fascin have been identified by 2-D gel electroNorthern analysis revealed that only four alleles, G409E, phoresis and Western blotting (YAMASHIR@MATSUMURA 116, 121, and S289N, contained significant levels of sn and MATSUMURA 1985; EDWARDS et al. 1995). We used transcripts (Figure 2A). The ovary-specific transcripts 2D gelelectrophoresis and Western blotting to analyze were missing from sn116.Western analysis of ovary exovary extracts from wild-type and sns2Xy~v females. Many tracts from these EMS sn alleles revealed that only two fascin isoforms wereresolved:however, the isoforms sn alleles, ~n""'~~'~' (fertile allele) and sn~s2Xy~" (sterile alpresent in wild typeovaries were also present in mutant sn S2XY.V ovaries (Figure 5). We also expressed mutant lele), contained detectablesinged protein (Figure 2B). S2XY.V fascin Egg chambers from sn';40y'~~ and sns2xy~v were further protein in the same PET 14b bacterial exanalyzed using immunofluorescence with antibodies to pression system (Novagen) used to express soluble wildDrosophila singed. Egg chambers from wild-type flies type protein (CANT et al. 1994). However, the mutant contained fascin in border cells, posterior follicle cells, protein was not soluble and could not be purified (data centripetal follicle cells and in the nurse cell cytoplasm not shown). (FIGURE 3A) (CANT et al. 1994). Egg chambers from Screen €or dominant suppressorsof the sterility pheboth s12 (;409b.' and snS289h' contained fascin in a pattern of s n s z ~ 9 ~ s: n s ~ ~ 9was s the best allele for further consistent with wild type (Figure 3, B and C). genetic analysis because it expressed a mutant protein The rapid phase of nurse cellcytoplasm transport and was sterile. To identify potential modifiers of fascin is an actindependent process. We analyzed the actin function, we performed a screen for dominantsuppresfilament networks in snc40y"and sn."Xy3"and compared sors of sn sterility.We mutagenized snSZXY'"males with 25 it with wild type (Figure 3, D-I). In egg chambersfrom mM EMS and then mated them to heterozygous snSzX9"' wild-type flies, a subcortical actin network is detected females (Figure 6). Out of 38,100 sns2Xy~v females, at thenurse cell membrane. Just before nurse cell rapid 28,100 were screened at 25" and the remaining 10,000 cytoplasm transport, cytoplasmic actin bundles form were screened at 18" to test for cold-sensitive suppresand extend from the plasma membrane to the nuclear sors. HomozygoussnSzRy,v female progeny that conmembrane (Figure 3D). As the nurse cell cytoplasm is tained mutagenized chromosomes were screened for transported into the oocyte, the oocyte grows, and the fertility by placing 20 females and 10 snS2XyN males in nurse cells regress (Figure 3G). sn""".'was a fertile alvials. Any F2 progeny in these vials must have been delele, but the cytoplasmic actin bundles appeared faint rived from a female sns2Xy,v that carried a dominant m
a-
- -
+
Fascin
of
253
Genetic Analysis
singed Ab
Rh Ph
F K X ' K I ~ ..?.-EMS sn mutants express fascin but cytoplasmic actin bundlcs arc disrupted. Egg chambcrs from wild type (A, D, and G), the fertile allele .m"~"J"'~~ (R,E, and H) and the sterile allele sn~s2"U'(C, F, and I) were stained either with sn monoclonal antibody (A-C) to detect fascin or rhodamine phalloidin (Rh Ph, D-I) to detect filamentous actin. Fascin is highly expressed in the nurse cell cytoplasm, border cells (BC), centripetal follicle cells (CFC), and posterior follicle cells (PFC) in wild type (A), sn f;-iOW.~(R),and snS2'".' (C). Stage 10B egg chambers from wild-type females (D) have a subcortical actin network just beneath the cell membrane and an extensive cytoplasmic actin bundle network (CAB). In wild type, all of the nurse cell cytoplasm is transported into the oocyte and the nurse cells regress (G). Although fascin is expressed in sn""9'.' (B),the cytoplasmic actin egg (H) shows that not all of the nurse cell cytoplasm is bundlcs appear sparse and disorganized (E). At stage 14, a sn'"' transported into the oocyte and the egg is only -75% the size of a wild-type stage 14 egg (G). Egg chambers from sn"-X"\females rescmble those seen in null singed alleles (CANT~t al. 1994); cytoplasmic actin bundles are nearly absent (F), the rapid phase of cytoplasm transport is disrupted, and mature eggs (I) are only half the size of wild-type eggs (G).
suppressor. Progeny were subsequently tested for fertility. Although no progeny were ever obtained from unmutagenized .snS2'"." females, -5% of the vials con@ining sn.s2xX"\' females with mutagenized chromosomes produced one to three progeny. These progeny were never fertile. One suppressor, Su(sn) was uncovered
from the 18" portion of the screen. sn.s2xy~v females carrying one copy of the Su(sn) mutation were fertile at 18" and at 25". Su(sn)not only was able to restore .wzs2xy~v fertility, but also suppressed thegnarled bristle morphology defect. sns2"'y\' females with one copyof Su(sn) had an intermediate kinked bristle phenotype, and
TABLE 1 EMSinduced singed alleles Phenotypes Fertility Allele sn~;JfJ'll~
sn.s2X'~.\'
sn"'
,sn.s',/"2x"
size
Mutation Egg
G409E S289N Unknown S289N S25l F,
Bristle Overall severity Intermediate Severe Nul1 Intermediate
Multiple bends Moderately gnarled Severely gnarled Subtlv bent
75% 50% 50% 80%
Fertile Sterile Sterile Fertile
K. Cant and L. Cooley
254 FLY SU
(24) W W T I G L I N G Q H K Y M T A E T F G K L N A N ~ S L K ~ Q L ~ L E P T .:.:ll:l:. : l : l l l . l l I:II.II.II :l:l;ll..:.: .I II:.. ..:II.I. 1 1 1 . 1 . ::I..:. (7) kykfglvnsagryltaekfggkvnasgatlkarqvwlleqeess.tisylkapsgnf1sadkngnvycsvedrteda
FLY (101) RGRFQISISEDGSGRWAbKCY..FLGGTPDKLVCT.AKTPGASEFWTVHLAARPQVNLRSIGRKRFAHL...SES 1:l.: .I I : I I I I I :1:...:.1:1. ..I..:I. I I I : I I , : I I l ::.:.:-l:lll .I: SU (83) d t g f e i e l q p d . . g k w a l k n v s h q r y l a c n g e e l i c s e s s t s n p s a n w t v q l h q r y a h l k t s e e g
.
:.
FLY (172) QDEIHVDANIPWGEDTLFTLEFRAEEGGRYALHTCNNKYLNANGKLQW~DCLFSAEYHGGH~~R~QYLSP :I.: 1 1 . : I I I . I . : I I : 1:l:l.. l.l::..:l.l....ll:. I. : :III.III..I..I:. SU (158) edsvwdelvpwgadstltlvylgk..gkygleafngkfvqtdgqla@aneqtqftliftliftsghlvlrdnngrhl~
:.
Sups251F
S289N
FLY (249) IGnKAVLKSRSSSVTRDELFSLEDSLPQASFIAGHNLR~Q~VT AN...QDEVGENETFQLEYDWSAHRWAL IIII...::I:.:.I I l l 1 I1:.l I :l.l:l1I II. IIIIII: ::I: SU (233) dsgtrvlksskpgltkanyfiledscpqgafefgg..kyaslkqgedvsfkllvdediedtetfqlefv.etdkyai
. ...... .......
... ..
......
...
FLY (323) RTTQDRYWCLSA GGGIQATGNRRC.ADALFELIWHGDGSLSFRGKFLATKRSGHLFATSESIEE1~ I..:.: : .I ::l!ll.ll.: .I. I.: ::I:: : . I I .ll::..: .Ill1 : SU (307) rvcdpkknsrdakfwktvaagiqangnskdqtdcqfsveyngnd.mhvrapggkyvsvrdnghlflqdspkd....
...
G409E FLY (392) F Y F Y L I N R P I L V L K C E ~ ~ G Y R T P G N L K L E C N W \ T Y E T I L V E R A Q K G L V I E G E S I S V D A D A P S D G F I:I I : I I I I I I I I . : I I I I I. .:.ll:..::.: I. : I.. :llll..::. I . I ::I: :.I SU (378) fifrllnrpklvlkcphgfvgmke.gkaevacnrsnfdvft~~keggyt.iqdscgkywscddssrlvlgeaa.gtf
:.
:..
.....................
FLY (470) FLELREPTRIC1RSQ.QGKYLGATKNGAFKLLDDGTDSATQWEF (512) 1:I::l .::.ll.: .I..: :...I I. : . : . . . . I I1 I SU (453) ffefhelskfairaesngmlikgeqsglftangsevskdtlwef (496) FIGURE 4.-A best fit alignment of Drosophila fascin (FLY) and sea urchin fascin (SU) is shown. The alignment of Drosophila PACKAGE of GCG (Version 8.1, 1995). Numbers and sea urchin fascin was generated with the best fit program in the WISCONSIN indicate amino acid number. Amino acids affected in sn(i109", sn"'""", and snsu~28gN are indicated in outlined letters that are underlined. Thesn'aoy"mutation lies in a highly conserved domain of 21 amino acids that is underlined with asterisks. Drosophila and sea urchin fascin are 35% identical, but in this 21 amino acid domain they are 71% identical.
sn S2R9.V males with one copy of Su(sn) had a very weak bristle defect visible only as subtle bends at the tips of the bristles (data not shown). We used the bristle suppression phenotype to map Su(sn). When Su(sn) males were mated with C(1) y f / Y females, all male progeny had the suppressed bristle phenotype indicating that theSu(sn)was on the Xchromosome. To map Su(sn) with respect to sns2R9N,we counted recombination events between Su(sn) and snS2X 9N in heterozygous Su(sn),sn""'"/ + females. There were 0/7968 recombination events between Su(sn) and SnS2R9,V . Therefore, the Su(sn) mutation appeared to be within 0.038 ( P < 0.05) map units of sn. Since Su(sn) was tightly linked to singed and also appeared to s u p press both the gnarledbristle and egg chamber phenotypes of sn"xyN,we tested whether Su(sn) was an intra-
genic suppressor of snS2R9N. We sequenced the sn open reading frame of sn"'", Su(sn) males and found that in addition to the original S289N mutation in sn, there was an additional C + T transition in codon 251 that changed a serine(TCC codon) tophenylalanine (TTC) (Figure 4). Flies containing the original S289N mutation and the suppressor S251F mutation in the same sn open reading frame will be referred to as ~n~'"@""~''.
basic
wild type
a
-
m
*
sn S289N FIGURE5.-Ovary extracts from wild type and from snszxyN were analyzed by2D electrophoresis and immunoblotting with singed monoclonal antibody 7C. Many isoforms of fascin are detected: however, the isoforms present in wild type are also present insns28y4vovaries. The isoform pattern is not affected by the S289N mutation.
Fertile? FIGURE 6.-EMS mutagenesis screen for dominant suppressors of snSZx9.'sterility. Mutagenized 70 snsZx9~'males were mated to heterozygous balanced snszXy." female virgins. Homozygous snS2XY.V female progeny containing mutagenized chromosomes were screened for fertility. Homozygous sn~"'q'.'' females should be sterile unless they carry a dominant suppressor of singed sterility. Any progeny obtained were further screened for fertility to determine if they inherited a suppressor.
2.55
Genetic Analysis of Fascin
FIGURE7.-Our dominant s u p pressor of snS2'"' sterility screen identified an intragenic mutation in singed referred to as snS1'~'28'1\: sn,.s"~,.\zs'l.Y containsan additional mutation, S251F, in the singed gene that can restore function to S2.YOV sn .S2II'I.Y females are sterile; nurse cell cytoplasmic actin bundles are nearly absent (A), the rapid phase of cytoplasm transport is disru ted,and a matureegg from snP,.y,. IS . only half the size of a wild type egg (D). Dominantly suppressed sn.%,p2""v,sn .$2W'.Y Se-
1
males are fertile and have sparse, disorganized cytoplasmic actin bundles (B). These few cytoplasmic bundles allow mature eggs to grow to -70% the size of wildtype eggs (D). Homozygous sn .s!,p2xY.\, sn .SU~,SZ
