UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The genetic and molecular analysis of the mfs(2)31 locus in Drosophila melanogaster : a novel suppressor… Whitehead, Ian P. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_spring_phd_whitehead_ian.pdf [ 10.83MB ]
JSON: 831-1.0086357.json
JSON-LD: 831-1.0086357-ld.json
RDF/XML (Pretty): 831-1.0086357-rdf.xml
RDF/JSON: 831-1.0086357-rdf.json
Turtle: 831-1.0086357-turtle.txt
N-Triples: 831-1.0086357-rdf-ntriples.txt
Original Record: 831-1.0086357-source.json
Full Text

Full Text

THE GENETIC AND MOLECULAR ANALYSIS OF THE mfs(2)31 LOCUS INDROSOPHILA MELANOGASTER: A NOVEL SUPPRESSOR OFPOSITION-EFFECT VARIEGATIONByIAN PAUL WHITEHEADB.Sc. (Hon.), University of British Columbia, 1987A THESIS SUBMITTED IN PARTIAL FUFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYINTHE FACULTY OF GRADUATE STUDIES(Department of Zoology)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAApril 1993© Ian Paul Whitehead, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of e7,40/00--7 The University of British ColumbiaVancouver, CanadaDate 27 /9,-,i/e3 DE-6 (2/88)ABSTRACTPosition-effect variegation (PEV) is the variable inactivationof a euchromatic gene which has been moved, by way of a chromosomalrearrangement, into a heterochromatic environment. Thetranscriptional repression at the variegating locus is thought tobe a consequence of inappropriate packaging of the euchromatin asheterochromatin. Second site mutations which modify the PEVphenotype (Su(var)s and E(var)s), identify loci which encode non-histone chromosomal proteins. Although the majority of mutationswhich modify PEV exhibit a dominant phenotype, rare recessivesu(var) mutations have been reported. This study describes a locuswhich is identified by one such recessive mutation, mfs(2)31.A cytogenetic analysis of subdivisions 31D-E was undertaken todetermine the precise location of the mfs(2)31 locus and to isolateadditional alleles. Five new deficiencies and 123 new lethalmutations were induced, allowing for the partitioning of 31D-E intosix cytological subintervals. The mfs(2)31 gene was localized todistal 31E in an interval containing nine lethal complementationgroups. Three new mfs(2)31 alleles were recovered, one of whichwas isolated in a screen for P element insertions.The new alleles of mfs(2)31 were used for a phenotypicanalysis of the locus.^Strong alleles that were lethal ashomozygotes died in the larval phase,^while weaker allelesexhibited the previously described bristle, sterility and su(var)phenotypes. Larval, pupal and adult functions were defined for theiilocus. Although no dominant phenotypes were observed, survivingheteroallelic combinations suppressed PEV in a variety ofvariegating backgrounds. When the weak, P-induced, allele wasoutcrossed in a dysgenic background, all the mfs(2)31 phenotypes,including suppression of PEV, were co-reverted. An in situhybridization to salivary gland polytene chromosomes revealed a Pelement in distal 31E of this strain.The P element, located in distal 31E, was cloned from thedysgenic mfs(2)31 allele. This element is responsible for thereduced transcription of a 1.2 kb message which is expressedthroughout development. Wild-type levels of transcription arerestored at this locus in revertants of the mfs(2)31 phenotype.The predicted amino acid sequence, as determined from a cDNAanalysis, reveals similarities with a mouse microtubule-associatedprotein and mammalian histone H1.iiiTABLE OF CONTENTSABSTRACT^ iiLIST OF TABLES^ ixLIST OF FIGURES xLIST OF BIOCHEMICAL ABBREVIATIONS^ xiiLIST OF GENETIC ABBREVIATIONS xiiiACKNOWLEDGEMENTS^ xvCHAPTER 1 - LITERATURE REVIEWA. General^ 1B. Chromatin Structure^ 2Histone Proteins 2Non-histone Chromosomal Proteins^ 3Heterochromatin and Euchromatin 6C. Position-effect Variegation^ 10The Phenomenon of Position-effect Variegation^10Position-effect Variegation and the Chromosomal^10BreakpointThe Timing of Position-effect Variegation^13Factors Which Modify Position-effect Variegation^14The Structural Basis for Position-effect^18VariegationDominant Modifiers of Position-effect Variegation^20Cloned Su(var) Loci^ 24Recessive Modifiers of Position-effect Variegation 26Summary^ 27ivCHAPTER 2 - CYTOGENETIC ANALYSIS OF SUBDIVISIONS 31D-EINTRODUCTION^ 29MATERIALS AND METHODSStocks^ 35Culture Conditions^ 37Cytology^ 37Genetic Screens^ 39Complementation 43Mapping by Recombination^ 43RESULTSCytology^ 45Screens 48Complementation^ 50Mapping by Recombination^ 53DISCUSSION^ 57CHAPTER 3 - GENETIC ANALYSIS OF THE mfs(2)31 LOCUSINTRODUCTION^ 62MATERIALS AND METHODSStocks^ 70Culture Conditions^ 71Complementation 71Fertility Tests^ 72Lethal Phases 72Pigment Assays^ 73Ovary Dissections 74Thick Sections of Ovaries^ 74Reversion Screen^ 75In situ Hybridization to Polytene Chromosomes^75RESULTSComplementation of mfs(2)31 With Su(var) Loci^79mfs(2)31 inter se Complementation^ 82Fertility of mfs(2)31 Alleles 85Studies on mfs(2)31 Oogenesis^ 88Lethal Phases for mfs(2)31 95mfs(2)31 Alleles Suppress Position-effect^97Variegationmfs2 has a P Element in 31E^ 100Reversion of mfs2^ 02DISCUSSION^ 108CHAPTER 4 - MOLECULAR ANALYSIS OF THE mfs(2)31 LOCUSINTRODUCTION^ 115MATERIALS AND METHODSStocks^ 118Recombinant Clones^ 118In situ Hybridizations 118DNA AnalysisIsolation of High Molecular Weight Genomic DNA^118Isolation of Plasmid and Cosmid DNA^119Small Scale Isolations^ 119Medium Scale Isolations 120Large Scale Isolations^ 121Isolation of Bacteriophage DNA 122Restriction Digests^ 123viGel Electrophoresis^ 123DNA Transfer From Agarose Gels to Nylon Membrane 124RNA AnalysisIsolation of Developmentally Staged Drosophila^124CulturesEmbryonic Collections^ 125Larval and Pupal Collections^ 125Isolation of Poly(A)'RNA^ 126Gel Electrophoresis of RNA 127RNA Transfer From Formaldehyde Gels to Nylon^128MembranesHybridization to Nylon Membranes^ 128Gel Purification of Probes 128Radiolabelling of Probes^ 128Filter Hybridization 129Library Construction and ScreeningConstruction of a Bacteriophage Library^129Screening Bacteriophage Libraries^ 131Preparation of Confluent Plates 131Plaque Lifts^ 131Screening Cosmid Libraries^ 132Subcloning Cosmid, Plasmid and Bacteriophage DNA 133Mapping DNA^ 133DNA SequencingSequencing Reactions^ 134Polyacrylamide Gel Electrophoresis^ 135Sequencing Strategy^ 135viiRESULTSCloning of a P Element From mfs 2^137Mapping X1^ 140Transcriptional Analysis of X1 and the mfs(2)31^148LocusSequencing mfs(2)31^ 156DISCUSSION^ 170LITERATURE CITED^ 176LIST OF TABLESTABLE PAGE1 Cytological limits and origins of chromosomalrearrangements362 Origin of previously described mutations in region 3831A-32A3 Summary of genetic screens 494 Mapping by recombination 545 Complementation with dominant Su(var)s 806 Complementation matrix for mfs(2)31 837 Fertility of mfs(2)31 hypomorphs 868 Lethal phases for mfs(2)31 alleles 969 Pigment assays 9810 Revertant description 10311 Revertant analysis 105LIST OF FIGURESFIGURE1^Screens for EMS induced lethal mutations2^Screen for P element induced lethal mutations3^Cytological map for division 31PAGE4041464 Cytogenetic map of region 31D-E 515 Recombination map for complementation groupslocated in cytogenetic interval #5566 Screen for revertants of mfs(2)31 767 Ovarian development in mfs(2)31 homozygotes 908 In situ hybridization to salivary gland polytenechromosomes of a mfs 2 b pr cn/+ strain1019 Reversion of su(var) and bristle phenotypes in ahybrid-dysgenesis induced revertant of mfs 2 .10710 mfs2 b pr cn/CyO genomic DNA hybridized with the 138900 by Hin dIII fragment from pn25.111 Bacteriophage X clones hybridized with the 900 by 139Hin dIII fragment from pn25.112 Localization of X1 to subdivision 31E by in situhybridization14113 Restriction maps for various fragments from X1 14214 Composite restriction map for X1 14315 Overlapping subcloned fragments from X1 hybridizedwith JT31 and JT3514516 Overlapping clones in subdivision 31E 14617 Digested 8.5 kb Eco RI fragment from X1 hybridizedwith pn25.114718 mfs(2)31 genomic DNA hybridized with the 4.0 kb 149Eco RI/Xba I fragment from X119 Transcriptional analysis of X1 15020^Putative mfs(2)31 cDNAs hybridized with pH1.5 and^152cDNA 621^Poly(A)' RNA from 0-6 hour embryos hybridized with^154cDNA 622^Poly(A)" RNA from the mfs(2)31 alleles hybridized^155with cDNA 623^Developmental profile of a putative mfs(2)31^157transcript24^The complete sequence of cDNA 6^ 15825^Similarities with the deduced protein sequence of^160mfs(2)3126^The relationship between the P element insert and^162the 5' end of cDNA 6 in mfs 2 .27^Partial sequence of the 1.0 Kb Hin dill fragment^163from the plasmid pH1.028^cos50 DNA hybridized with pH1.5^ 16429^Overlapping fragments from X1 hybridized with^165cos5030^The complete genomic sequence of the region^167encompassing the gene represented by cDNA 631^Potential upstream regulatory sequences of mfs(2)31^168xiLIST OF BIOCHEMICAL ABBREVIATIONSAPS^ammonium persulphateby base pairsBCIP^5-bromo-4-chloro-3-indoly1 phosphateDTT dithiothreitolEDTA^ethylenediaminetetraacetic acidEMS ethyl methanesulfonatekb^kilobase pairsX-dil^100 mM NaC1, 10 mM Tris-HC1 (pH 7.4), 10 mM MgC1 2LM 10 g/1 tryptone, 5 g/1 yeast extract, 2 g/1 MgC1 2 ,5 g/1 NaC1MOPS^3-[N-morpholino]propanesulfonic acidNBT nitro-blue tetrazoliumnt^nucleotidesPEG polyethyleneglycolpfu^plaque forming unitsPVP polyvinylpyrrolidineSSC^standard saline citrate; 0.15 M NaC1, 0.015 M sodiumcitrateTE^10 mM Tris-HC1 (pH 8.0), 1 mM EDTATEMED^N,N,N',N'-tetramethylethylenediamineX-gal^5-bromo-4-chloro-3-indolyl-beta-D-galactosidexiiGENETIC ABBREVIATIONSSTRAINSOR^Oregon-R; stock derived from wild flies collected in1925 at Roseburg, Oregon (Lindsley and Grell, 1968).Canton-S Canton-Special; stock derived from wild flies collectedin Canton, Ohio (Lindsley and Grell, 1968).CHROMOSOMESBirm2CyOA2-3TM6win4A 2nd chromosome from the M' strain, Birmingham. Itbears 17 defective P elements (Robertson et al., 1988).Curly derivative of Oster; In(2LR)0, Cy dp 1 Vz pr cn 2 ;balanced 2nd chromosome (Lindsley and Zimm, 1992).2-3](99B); 3rd chromosome with a P insertion whichprovides transposase activity (Robertson et al., 1988)In(3LR)TM6, Hh' ss P88 b.X34e Ubx'15 e; balanced 3rdchromosome (Lindsley and Zimm, 1992)./n(/)w 4 ; X chromosome inversion which variegates forthe white gene. The euchromatic breakpoint is distal towhite and the heterochromatic breakpoint is distal to theribosomal RNA genes (Hilliker et al., 1980).win 51b^In(1) Wm51b; X chromosome inversion which variegates forthe white gene. The euchromatic breakpoint is distal towhite and the heterochromatic breakpoint is proximal tothe ribosomal RNA genes (Hilliker et al., 1980)./n(/)re7; X chromosome inversion which variegates forthe white gene. The euchromatic breakpoint is proximalto white and the heterochromatic breakpoint is proximalto the ribosomal RNA genes (Tartof et al., 1984).MARKERSbBccnblack (2-48.5); black pigment on body (recessive.Black cells (2-80.6); black cells appear in 1st instarlarvae (dominant); homozygous lethal.cinnabar (2-57.5); bright red eye color (recessive).Cy^Curly (2-6.1); wings curled upwards (dominant);homozygous lethal.Gl aJPrry506SbSpTftGlazed (unknown); reduced gleaming eyes (dominant);homozygous lethal.Jammed (2-41.0); narrow, fluid-filled wings (dominant);homozygous lethal.purple (2-54.5); purplish, ruby eye color (recessive).rosy (3-52.0); reddish-brown eye color (recessive).Stubble (3-58.22); short, thick bristles (dominant);homozygous lethal.Sternpleural (2-22.0); sternopleural bristles increasedin number (dominant); homozygous lethal.Tufted (2-102); tufts of bristles on mesothorax(dominant); homozygous lethal.xivACKNOWLEDGEMENTSI would like to acknowledge and thank my supervisor, TomGrigliatti, for his encouragement, support and advice throughoutthe course of this work. I would also like to thank my supervisorycommittee for critical reviews of this thesis: Jim Berger, DaveHolm, Don Moerman and George Spiegelman.There have been a number of people who have passed through thelab with whom I have had the pleasure of working and with whom Ihave developed lasting friendships. I would like to thank DonSinclair, Nigel Clegg, Amy Hedrick, Jo-ann Brock, Mike 0-Grady,Murray Richter and Gunter Reuter for many fond memories of graduateschool.On a more personal note, I would like to thank Fiona, Paul,Mum, Dad, Gwen and, of course, Damien. Your contributions to thecompletion of this work far exceeded all others.CHAPTER 1: LITERATURE REVIEWA. GENERALDuring a typical cell cycle, a eukaryotic nucleus willtranscribe, replicate, repair and mobilize between 10 7 and 10' byof DNA. In order to ensure the cell's survival, these events mustbe orderly and occur with a high degree of fidelity. To facilitatethe execution of these complex processes, DNA is packaged in ahierarchical manner. DNA packaging is dynamic and heterogeneous;chromosomes acquire different states of organization as theyproceed through the cell cycle, and often exhibit several states ofcompaction at any one time.^The manipulation of chromosomalarchitecture^is undoubtably the result of complex metabolicprocesses, yet little is known about the enzymatic activitiesinvolved or the substrates upon which they act.A change in DNA packaging can often be visualized at themicroscopic level as an alteration in the morphology and stainingproperties of chromatin. Chromatin is both a cytological and abiochemical term used to describe DNA and associated proteins, thesubstrates from which chromosomes are built. Changes in chromatinorganization are the result of altered structural relationshipsbetween its component parts, both protein and DNA. Anunderstanding of the character and function of the histone and non-histone proteins, which constitute 2/3 of the chromatin mass, wouldbe prerequisite to an understanding of chromosomal and nucleararchitecture.1B. CHROMATIN STRUCTUREHistone proteins Apart from the DNA itself, the most extensively studiedcomponent of chromatin is the histones. These highly conserved andubiquitous proteins comprise approximately 1/3 the mass ofchromatin and have an almost universal role in eukaryotic DNApackaging. At the simplest level of chromatin organization, DNAwinds around histone octamers forming a fiber 10 nm in diameter.The octamers are spaced 30 by apart and consist of a tetramercontaining two H3 and two H4 molecules and two H2A-H2B dimers(reviewed by McGhee and Felsenfeld, 1980). The so-called 10 nmfiber is further organized by histone H1 into a more compactstructure, 30 nm in diameter . Although the structure of the 30 nmfiber is uncertain, most evidence points to a simple solenoidalconformation with six nucleosomes per helical turn (reviewed byFelsenfeld and McGhee, 1986). This structure is compact, stableunder physiological conditions, and presents a considerablebarrier, in vitro, to the passage of RNA polymerase. Higher orderlevels of chromatin structure give DNA a maximum order ofcompaction greater than 10,000. Although little is known of thenature of these packing structures, most evidence suggests that thechromatin fiber is organized into discreet domains (ie. loops)whose integrity is maintained by proteins other than the histones(reviewed by Jackson, 1991).In addition to fulfilling an important structural role withinchromatin, histones participate directly in the regulation of gene2expression, either by blocking access to regulatory sequences or byinteracting with proteins involved in transcriptional regulation(reviewed by Grunstein, 1990). Such selective and localizedactivity on the part of a ubiquitous and highly conserved family ofproteins is achieved through an array of enzyme-catalyzed post-translational modifications. All core histones can be modified byacetylation, phosphorylation, ADP-ribosylation and ubiquitinizationof specific amino acid residues (Wu et al., 1984). Many of thesemodifications occur in the amino terminal domains which falloutside the core DNA and thus are ideally situated to interact withother DNA sequences and non-histone proteins (Hill and Thomas,1990). Of these post-translational modifications, acetylation hasbeen most commonly linked to transcriptional regulation (Ip et al.,1988; Hebbes et al., 1988). All four core histones undergo cyclesof acetylation and deacetylation which are correlated with changesin transcriptional competence and with the cell cycle (reviewed byTurner, 1991). Although the mechanism of transcriptionalregulation associated with histone acetylation remains unclear,functional differences between histones appear to depend as much onwhich sites are acetylated as on the overall level of acetylation(Johnson et al., 1990; Megee et al., 1990).Non-histone chromosomal proteins In addition to the DNA and histones, non-histone chromosomalproteins (NHPs) account for 1/3 of the chromatin mass. Of the NHPswhich are known to influence chromatin structure, the moderately3abundant high mobility group (HMG) are the best characterized(reviewed by Johns, 1983). The term HMG is operationally definedby the biochemical properties which allow these proteins to beisolated. Like the histone proteins, the HMGs are highlyconserved, they undergo a wide spectrum of post-translationalcovalent modifications and alterations in their stoichiometry havebeen correlated with cellular proliferation and differentiationevents (Johns, 1983; Weisbrod et al., 1980; Dorbic and Wittig,1987).Recently, monoclonal antibodies have become increasinglypopular as a tool for the dissection and analysis of nuclearprotein fractions. In Drosophila melanogaster, this approach hasidentified several proteins which are probably involved inchromatin condensation, the best characterized of which are D1, HP1and BJ1.D1 is a highly abundant 50 kDa protein which binds to AT richregions of DNA such as the highly-repetitive satellite sequences(Rodriguez-Alfageme et al., 1980; Levinger and Varshaysky, 1982).It shares many biochemical properties with the HMG proteins andshows limited sequence similarity with HMG I, a satellite DNAbinding protein from African Green Monkey (Strauss and Varshaysky,1984; Ashley et al., 1989). Although the physiological role of D1remains unclear, since the protein specifically associates withsatellite DNA it is plausible that it contributes to the highdegree of compaction of these heterochromatic sequences.HP1 is a 19 kDa protein isolated from a fraction of nuclear4proteins which have tight associations with DNA (James and Elgin,1986). Immunofluorescence staining of salivary gland polytenechromosomes with an antibody to HP1 shows specific binding to 0-heterochromatin, a banded portion of the heterochromatic fourthchromosome, and several discrete regions within euchromatin (Jamesand Elgin, 1986; James et al., 1989). Although the physiologicalrole of HP1 is unknown, the protein sequence does not contain anypreviously described DNA binding motifs. A detailed description ofthe HP1 protein and its associated locus are provided elsewhere inthis survey (see Cloned Su(var) loci).BJ1 is a 68 kDa nuclear protein which associates withnucleosomes and is released from chromatin by agents whichintercalate with DNA (Frasch, 1991). When hybridized to salivarygland chromosomes, an antibody to BJ1 shows no site preference,associating with all chromosome bands. The BJ1 protein sequenceshows strong homologies with the human gene Regulator of ChromatinmRNA metabolism andproposed that BJ1maintain a higherexpression (Frasch,cell cycleinvolved inmaintenance of nuclear structure. It has beenbinds to chromatin to either establish ororder structure required for regulated gene1991).Condensation (RCC1), a gene which functions inregulation, and the Yeast gene SRM1/PRP20, which isA third approach to identifying and analyzing NHPs which areinvolved in chromatin condensation employs genetic assays. Geneticapproaches have several advantages over the more traditionalbiochemical and structural approaches. A sensitive genetic assay5can be used to identify proteins present in very low abundance orwith very transient or weak associations with chromatin. Inaddition, the isolation of mutations in the genes encoding theseproteins allows for an analysis of their role in development andcan reveal unexpected associations and interactions with othercellular components. One such assay, position-effect variegation(PEV), may prove to be a valuable genetic tool for identifyinggenes which affect two cytologically differentiable states ofchromatin compaction, euchromatin and heterochromatin.Heterochromatin and euchromatin Eukaryotic chromosomes achieve their greatest level ofcompaction and appear homogeneous in morphology in late prophase asthey prepare for segregation. By telophase, the homogeneitydisappears with the majority of the DNA becoming loosely packed anddispersed, a state referred to as euchromatin, while a significantportion of the genome, the heterochromatin, retains a high level ofcompaction. In Drosophila melanogaster, this heterogeneity can beobserved in the neural ganglia.Heterochromatin, the portion of the DNA that remains compactedduring interphase, is an almost universal feature of the eukaryoticnucleus. Its distribution is non-random, and is most oftenassociated with centromeres and telomeres, although whole arm andinterstitial heterochromatization is not uncommon. Apart fromcytological morphology, heterochromatin differs from euchromatin inthat it generally replicates later in the cell cycle (Barigozzi et6al., 1966), exhibits lower levels of genetic activity (Hilliker,1976; Hilliker et al., 1980; Marchant and Holm, 1988) and itundergoes little or no meiotic recombination (Roberts, 1965;Schalet and Lefevre, 1976).In Drosophila, heterochromatin constitutes nearly 1/2 of theacrocentric X chromosome, the proximal 1/4 of the metacentricsecond and third chromosomes, all of the Y, and probably most ofthe small chromosome 4. It is first observed very early inDrosophila development, between stages 10 and 13 of the syncytialblastoderm (Foe and Alberts, 1985), its appearance being correlatedwith the beginning of transcriptional competence for the zygoticgenome (Edgar and Schubiger, 1986).Drosophila heterochromatin is divided into two classes, alpha(a) and beta (0), which are distinguishable both at a cytological(Heitz, 1934) and a biochemical level (Gall et al., 1971; Spradlingand Rubin, 1981; Healy et al., 1988; Devlin et al., 1990). Whereasin most eukaryotic tissues examined, euchromatin appears diffuseand is dispersed throughout the nucleus, a notable exception can befound in the salivary glands of dipterans. Here the chromosomes gothrough numerous rounds of endoreduplication achieving a ploidy ofgreater than 1000. The DNA from each round of replication remainsassociated with the previous copy, thus producing morphologicallydistinct giant chromosomes which can be readily seen using lightmicroscopy. Each chromosome has a unique banding pattern providinga valuable tool for the Drosophila geneticist (Lefevre, 1974;1976). It is in salivary gland chromosome preparations that a- and713-heterochromatin can be cytologically distinguished (Heitz, 1934) .a-heterochromatin is the chromatin closest to the centromere. Itis underreplicated by a factor of 10 relative to euchromatin inpolytene preparations, appearing thin and unbanded (Rudkin, 1969;Gall et al., 1971). The a-heterochromatin from the variousDrosophila chromosomes associates during interphase to form thechromocenter. At the biochemical level, a-heterochromatin is richin highly repetitive satellite DNA with few, if any uniquesequences (Gall et al., 1971).D-heterochromatin is the portion of the Drosophila chromosomewhich connects a-heterochromatin to euchromatin (Heitz, 1934). Itis not underreplicated in the salivary glands but it is verycompact, dark staining and unbanded (Rudkin, 1969). The DNAcontent of 0-heterochromatin differs from a-heterochromatin inthat it contains more unique sequence and middle-repetitive DNA(Spradling and Rubin, 1981; Healy et al., 1988; Devlin et al.,1990). Unlike a-heterochromatin, 13-heterochromatin is known tocontain actively transcribed genes (Biessmann et al., 1981; Mikloset al., 1984; Devlin et al., 1990) but at a considerably lowerfrequency than euchromatin (Hilliker et al., 1980; Marchant andHolm, 1988). There is no known functional basis for distinguishingbetween a- and P-heterochromatin.Although long thought to be genetically inert, studies onDrosophila heterochromatin, have revealed an unexpectedly highlevel of genetic activity (Hilliker et al., 1980; Marchant andHolm, 1988). Utilizing heterochromatic deficiencies, generated by8detaching compound autosomes, complementation maps have beenelaborated for essential loci in the centromeric heterochromatin ofchromosomes 2 (Hilliker and Holm, 1977) and 3 (Marchant and Holm,1988). A level of genetic activity approximately 100-fold less (ona per kb basis) than a comparably sized euchromatic segment isobserved. Although there is no genetic principle linking the geneslocated within heterochromatin, these genes can act late indevelopment, well after heterochromatin becomes established as acytological entity (Hilliker, 1976). The light gene which islocated in the centromeric heterochromatin of chromosomal arm 2L istranscribed throughout development suggesting that heterochromatindoes not necessarily form a transcriptional block to sequenceslocated within it (Devlin et al., 1990).It is plausible that NHPs are partially responsible formaintaining the difference in chromatin compaction that existsbetween euchromatin and heterochromatin. Proteins such as HP1 andD1, which are known to associate specifically with heterochromatin,may perform such a function. It may be possible to identify otherproteins involved in this process through genetic assays such asposition-effect variegation (PEV). It is notable that the genewhich encodes HP1, Su(var)205, was initially identified as amodifier of PEV (Sinclair et al., 1983).9C. POSITION-EFFECT VARIEGATIONThe phenomenon of position-effect variegation Position-effect variegation (PEV) is the variable inactivationof a gene caused by its repositioning in the genome. Thephenomenon differs from more classic position effects in that therelocated gene is inactivated in only a subset of the cells inwhich it is normally expressed. Thus, if the affected gene acts ina cell autonomous manner, the inactivation can often be observed asa mosaic phenotype. First described by Muller (1930) in Drosophilamelanogaster, PEV has since been observed in vertebrates(Cattanach, 1974), lower eukaryotes (Clutterbuck and Spathas, 1984)and plants (Catcheside, 1938; 1947). Although it has beendescribed in representative species from all kingdoms, thephenomenon has been most extensively studied in the Drosophilids(reviewed by Baker, 1968; Spofford, 1976; Eissenberg, 1989;Spradling and Karpen, 1990; Henikoff, 1990; Grigliatti, 1992), abias which is reflected in the following survey. Unless otherwiseindicated, all experiments discussed used Drosophila melanogasteras the experimental organism.Position-effect variegation and the chromosomal breakpoint The earliest described variegated position effects involveeuchromatic genes which have been juxtaposed, by way of chromosomalrearrangements, with broken segments of heterochromatin. Suchrearrangements are typical of variegating strains and there isconsiderable evidence that inactivation is the result of an1 0interaction of the affected gene with the adjacent heterochromatin.Although the chromosomal breakpoint involved is typically in a-heterochromatin (Tartof et al., 1984), the source of theheterochromatin does not appear to be an important factor(Spofford, 1976). Variegating position effects have been describedfor virtually all autosomal genes, two exceptions being ebony(Brosseau, 1970) and the bithorax complex (E.B. Lewis, cited inHenikoff, 1990).Although PEV typically involves an association between aeuchromatic gene and a broken segment of heterochromatin, similareffects have been described for heterochromatic genes which arerelocated next to euchromatin. The light gene and cubitusinterruptus are located in 0-heterochromatin of the second andfourth chromosomes respectively and both exhibit variegation whenmoved into a euchromatic environment (Stern and Kodani, 1955;Hessler, 1958). Interestingly, only breakpoints in centric anddistal euchromatin can induce light variegation (Hearn et al.,1991).There is considerable evidence that PEV is the result of aninteraction between the affected gene and the adjacent brokensegment of heterochromatin. The first evidence for thisinteraction came from experiments in which this relationship wasdisturbed. It has been well documented (Judd, 1955) that PEV onlyoccurs when a gene is located cis to the broken heterochromatin.A possible rare exception to this rule may involve the bwvvariegating rearrangement (Henikoff, 1981). Complete wild type1 1function can be restored to a variegating gene if it is removedfrom the broken segment of heterochromatin by recombination (Judd,1955). Similarly, wild-type function is restored to a variegatinggene if it is removed from the broken segment by a secondarychromosomal rearrangement (Hinton and Goldsmith, 1950; Reuter etal., 1985). Restoration of full gene expression does not requirethat the gene be returned to its normal euchromatic location, anobservation which implies that variegation is not caused by movinga gene from its native environment.A second observation that implicates the heterochromatin inthe variegating phenotype is that more than one gene can variegatein the same rearrangement and that the probability of a genevariegating is a function of its distance from the heterochromatin.This polarity of effect is most evident in the T(1;4) wm258-21variegating rearrangement (Demerec and Slizynska, 1937; cited inCohen, 1962), a strain which variegates for the white gene(abnormal eye color) and the roughest gene (disorganized eyefacets), both of which are expressed in a cell autonomous manner.The two genes are tightly linked genetically with roughest beingclosest to the broken heterochromatin. An analysis of thevariegated eye reveals that white clones are always smaller thanand completely contained within roughest clones. Thus, in no caseis the distal white gene inactivated without the more proximalroughest gene also being inactivated. A similar spreading effecthas been described for heterochromatic genes in Drosophila hydei(Hess, 1970).12To date, there have been no substantiated exceptions to thepolarity of variegation which appears to emanate from the brokenpiece of heterochromatin. Spreading effects have been describedover distances of 80 polytene chromosome bands (1500 kb) from thebroken^heterochromatin^(Schultz,^1950)^although^typically^thedistance is much shorter. The extent of spreading appears to bedetermined^by qualities of^both the^euchromatin^and^theheterochromatin involved in^the variegating^rearrangement(Spofford, 1976).A final observation linking variegation to the brokenheterochromatin comes from cytological studies. In polytenechromosome preparations from variegating strains, the euchromatinadjacent to the broken heterochromatin exhibits a morphologyreminiscent of 0-heterochromatin i.e. compact, dark-staining andunbanded (Hartmann-Goldstein, 1967). In the variegatingrearrangement T(1;4)wirr258-21, the white gene located at band 3C2 iscloser to the broken heterochromatin than the notch gene at band3C7. The relative visibilities of these two bands in thevariegating rearrangement is subject to the same polarity observedat the phenotypic level (Hartmann-Goldstein, 1967). Thus, eitherboth bands are seen, neither are seen or only 3C7 is seen. In nocase is 3C2 visible when 3C7 is not.The timing of position-effect variegation Clonal^analyses of variegating rearrangements havedemonstrated that the transcriptional competence of a variegating13gene is determined early in development, and once decided, ispropagated clonally (Janning, 1970). These conclusions are basedon the observation that variegation is generally characterized bylarge spot clones which follow the same boundaries as celllineages. In a number of variegating rearrangements, the timing ofthe determinative event has been calculated with a moderate degreeof precision. The rearrangement In(l)scs1 variegates for the rDNAcistrons. Newly hatched larvae with the genotype In(1) scS1 /0contain 14% less rDNA than their wild-type controls (Puckett andSnyder, cited in Spofford, 1976). Variegation of the yellow genein In(1)sc8 occurs as early as cellular blastoderm as does roughestvariegation in In(l)rst 3 . Variegation prior to early first instarlarvae has been detected for the peach gene in Drosophila virilis(in T(Y;5)pem 1 ; Baker, 1967) as well as a number or eye colormutations in Drosophila melanogaster (Baker, 1967; Janning, 1970).Although the bulk of the evidence points to an early determinativeevent, there are several variegating strains such as /n(/)W'' b andIn(1)4Pmc which show a very fine grained mosaic phenotype (Tartof,1984) indicating that in some cases the decision is made late orthat an earlier decision can be reversed. Regardless of the timeat which variegation is initiated, all of these studies demonstratethe clonal nature of the phenomenon.Factors which modify position-effect variegation A number of genetic studies of the PEV phenomenon havedescribed environmental, biochemical and genetic factors which have14a modifying effect on the variegating phenotype. The nature ofthese factors has fuelled many theories on the underlyingstructural mechanism of variegation. The environmental factorhaving the greatest impact on the extent of expression of avariegating locus is temperature (Gowen and Gay, 1935). Culturesraised at high temperatures (29°) have a lower proportion of cellsin which the variegating gene is inactivated than do culturesraised at lower temperatures (22°). Although the effect ismeasurable and reproducible, it is not always striking. Beingpoikilotherms, drosopholids develop more rapidly at 29° than at 22°.The extent of variegation is correlated with length ofdevelopmental time (see below), which could explain the temperatureeffect. It has also been proposed that variegation involves theassembly of macromolecular complexes and that in some instanceshigh temperatures do not favor the formation of such complexes(Zuckerkandl, 1974). A third explanation would be that at 29°flies are physiologically stressed and individuals showing strongvariegation would be selected against. At this time there is noevidence to support or refute any of these three hypotheses.It has been known for some time that flies developed incrowded cultures show stronger variegation than those developed innon-crowded cultures (Hinton, 1949). Because of competition fornutrients, flies raised in crowded cultures have a longerdevelopmental time and this has been invoked as the explanation forthe enhancement of variegation. Proponents of the hypothesis thatthe extent of variegation reflects the ability of a cell to15assemble heterochromatin argue that a longer cell cycle associatedwith prolonged development would allow more time for assembly tooccur. Regardless of the underlying mechanism, the evidence linkingstrength of variegation to developmental time is somewhatequivocal. Flies reared at pH 2.6 which results in a prolongeddevelopmental time show enhanced variegation (Michailidis et al.,1988). Some mutations which cause delayed development also showenhancement of PEV; however, there are some mutations which don't(Michailidis et al., 1988).There are chemicals which have been shown to modify PEV andalso influence developmental time. A number of agents which delaydevelopment by interfering with DNA synthesis have been correlatedwith enhancement of PEV (Schultz, 1956). However, proprionate andbutyrate, both of which prolong developmental time, suppressvariegation of the white gene in Drosophila melanogaster (Mottus etal., 1980; Rushlow et al., 1984). Butyrate is a potent inhibitorof histone deacetylase (Candido et al., 1978) and is known toaffect chromatin compaction (Annunziato et al., 1988). Although itis tempting to conclude that butyrate modifies variegation throughan alteration in chromatin structure, this chemical has been shownto have a wide spectrum of effects on other cellular processeswhich may be responsible for the modification (Boffa, 1981;Christman et al., 1980).In addition to being sensitive to environmental and chemicalinfluences, the PEV phenomenon can be modified by a number of welldescribed genetic factors. Strength of variegation is very16sensitive to the ploidy of the Y chromosome. Males with extra Ychromosomes (XYY males) exhibit a suppressed phenotype (Gowen andGay, 1934) while the absence of a Y chromosome (X0 males) enhancesvariegation (Spofford, 1976). The effect appears to beattributable to the amount of Y chromosome present rather than anyspecific sites (Dimitri and Pisano, 1989). Zuckerkandl (1974) andothers have proposed that PEV is sensitive to intracellularavailability of structural components of heterochromatin and thatthe Y chromosome acts as a sink for these components. The entireY chromosome in Drosophila melanogaster is packaged asheterochromatin. Zuckerkandl's proposal finds support in theobservation that deficiencies and duplications of autosomalheterochromatin affect variegation in the same way as deficienciesand duplications of the Y chromosome (Spofford, 1976).The proposal that PEV is sensitive to intracellular levels ofthe structural components of heterochromatin was examined bytesting the sensitivity of PEV to the dosage of the histone genecluster (Khesin and Leibovitch, 1978; Moore et al., 1983).Strains which are haploid for the histone gene cluster exhibitstrong suppression of PEV in rearrangements which variegate for thewhite gene (Khesin and Leibovitch, 1979; Moore et al., 1983).However, in similar studies, no effect was observed at the rosylocus (Rushlow and Chovnick, 1984). It has been proposed thatchromatin formation is sensitive to the availability of a number ofchromatin components and that heterochromatin, because itreplicates late, is more sensitive (Moore et al., 1983).17Since PEV is sensitive to the dosage of histone proteins, thissensitivity may extend to the dosage of genes encoding otherchromosomal proteins involved in chromatin condensation. A largenumber of single-site modifiers of the PEV phenotype have beenisolated, which may identify proteins of this class. These aredescribed in detail in the section titled Dominant modifiers of PEV.The structural basis for position-effect variegation Variegation is not associated with somatic loss of thevariegating locus, but rather, a reduction in the accumulation ofnascent transcripts. This has been observed for variegatingalleles of the Hsp70 gene (Henikoff, 1981), the rosy gene (Rushlowet al., 1984), the Sgs4 gene (Kornher and Kauffman, 1986) and thebrown gene (Henikoff and Dreesen, 1989). It is not yet clearwhether the reduction in nascent transcription at the variegatinglocus is due to under-replication, reduced accessibility of thelocus to transcriptional regulators or transcriptional interferencefrom a promoter located within the adjacent heterochromatin.If unique probes for variegating genes are hybridized tosalivary gland DNA there is no evidence of underreplication for thewhite gene (Hayashi et al., 1990), the Hsp70 gene (Henikoff, 1981)or the rosy gene (Rushlow et al., 1984). However, underreplicationhas been detected at the Sgs4 locus in T(1;4)wm258-21 (Koernher andKauffman, 1986) and at the yellow locus in Dp(1;f)1187 (Karpen andSpradling, 1990). Although these latter studies implicate18underreplication as one mechanism for variegation it cannot beinvoked to explain variegation in diploid tissue. The white genevariegates in pigment cells of the eye while the nod genevariegates in germline cells, two tissues which are neitherpolytenized nor polyploid (Karpen and Spradling, 1990, Zhang andHawley, 1990).Another plausible explanation for the reduction in nascenttranscription associated with a variegating gene would be aninaccessibility of the locus to transcriptional regulators. Anumber of groups have described a positive correlation between thestrength of a variegating rearrangement and the number of salivarygland cells in which the chromosomal segment containing theaffected gene adopts an altered cytological morphology (Hartmann-Goldstein, 1967; Henikoff, 1981; Reuter et al., 1982; Zhimulev etal., 1986; Hayashi et al., 1990). The affected segment becomesdark-staining, compact and unbanded, features reminiscent of 0-heterochromatin. In a series of in situ hybridizations to salivarygland chromosomes, the white gene has been shown to be lessaccessible to DNA probes in variegating rearrangements than in non-variegating strains (Hayashi et al., 1990).Since variegating genes are invariably associated with brokensegments of heterochromatin, a simple interpretation of theobservations described above would be that variegation is theresult of a euchromatic segment of DNA taking on characteristics ofheterochromatin. In diploid tissue, variegation would be caused byovercompaction of the DNA while in polytenized tissue it would be19caused by underreplication. In this regard, it is interesting thateuchromatic sequences juxtaposed to heterochromatin become latereplicating (Prokovyeva-Belgovskaya, 1947; Ananiev and Gvozdev,1974), a feature normally associated with heterochromatin.An intriguing, alternative explanation for variegatingphenotypes is that transcription from a promoter within theheterochromatin (ie. in a transposable element) is interfering withtranscription at the affected euchromatic locus (Spradling andKarpen, 1990). The heterochromatic breakpoints of severalvariegating rearrangements are closely associated with satelliteand middle-repetitive DNA sequences (Tartof et al., 1984) and ithas been proposed that transcription may initiate within thesesequences. In several rearrangements which completely restorewild-type function to the variegating loci, these elements arestill associated with the gene implying that variegation originatesfrom deeper within the heterochromatin. However, Reuter et al.(1985) show that in 48 of 51 wild-type revertants of a variegatingrearrangement, an affected gene can still be made to variegate inthe presence of a strong Enhancer of PEV (see Dominant modifiers of PEV). This suggests that variegation can in fact originate fromrelatively small sequences, adjacent to the chromosomal breakpoint.The revertants characterized by Tartof (1984), do not respond tostrong E(var) loci (T. Grigliatti, personal communication).Dominant Modifiers of position effect variegation Genetic modifiers of PEV either reduce (Su(var)s) or increase20(E(var)s) the number of cells in which the variegating gene isinactivated. A large number of single-site genetic modifiers ofPEV have been described (Reuter and Wolff, 1981; Sinclair et al.,1983; Sinclair et al., 1989; Locke et a/., 1988; Wustmann et al.,1989). The majority of these have been identified in segmentalaneuploidy studies (Locke et al., 1988; Wustmann et al., 1989).Locke et al. (1988), use 12 chromosomal rearrangements to definefour independent loci and from this extrapolate to predict theexistence of 20-30 dosage-sensitive modifiers of PEV in theDrosophila genome. In a more extensive study, Wustmann et a/.(1989) describe 38 independent loci in an area covering 30% of theautosomal complement. They predict the existence of 120-150 suchloci, a number closely approximated by several other studiesconducted on a smaller scale (Henikoff, 1979; Reuter et al., 1987;Szidonya and Reuter, 1988). Altogether, the aneuploidy studiesidentified 44 dosage dependent modifying loci which can beconveniently divided into four classes: haplo suppressor/triploenhancer, haplo enhancer/triplo suppressor, haplo enhancer andhaplo suppressor. The reciprocal haplo/triplo phenotypes of theloci which fall into the first two categories suggests that theymay play an important role in the chromatin assembly process.A model has been proposed to explain the sensitivity of PEV tothe dosage of a number of single-site modifiers (Locke et al.,1988). Arguing that heterochromatin formation involves the self-assembly of multimeric complexes and that Su(var) loci identifycomponents of these complexes, Locke et al. (1988) point out that21the law of mass action dictates that the dose of a single componentcan effect the rate of assembly exponentially. If none of thecomponents are rate-limiting, the rate would be determined by thetotal number of elements in the complex. Although an attractivehypothesis, it may only apply to the limited number of modifierswhich show reciprocal haplo/triplo effects.In addition to modifiers of PEV identified by segmentalaneuploidy, two large collections of ethylmethane sulfonate (EMS)induced Su(var) mutations have been described. One collectionconsists of 12 mutations on the second chromosome and elevenmutations on the third chromosome (Reuter and Wolff, 1981) whilethe second collection identifies 16 on the second and 33 on thethird (Sinclair et al., 1983). Smaller collections of E(var) locihave also been described (Reuter and Wolff, 1981; Sinclair et al.,1989). The location of these loci correspond well with the lociidentified in the aneuploidy studies and their phenotypes indicatethat the majority of mutations are amorphs or strong hypomorphs.The elaboration of a complementation map comprised of both Su(var)collections is the focus of a collaborative effort between the twogroups (R. Mottus, personal communication).Typically, Su(var) mutations exhibit a dominant phenotype,often restoring nearly full activity to a variegating gene. Theyexert their influence over a broad range of variegating genesshowing no preference for particular types of rearrangements(Hayashi et al., 1990; Reuter et a/., 1986; Sinclair et al., 1983;1991). Their suppressing ability is often sex-specific (Hayashi et22al., 1990; Reuter et a/., 1986; Sinclair et al., 1991) and allele-specific, some alleles showing stronger suppression than others.Not all of the suppressor genes are essential. Reuter et al.(1986) report that all of their third chromosome suppressors areeither homozygous lethal or sterile while Sinclair et al. (1983),note that all of their third chromosome suppressors are homozygousviable. These two collections may not comprise the same loci, oralternatively, the results may reflect differences in the strengthof particular alleles. The germ-line requirement for suppressorgenes has been examined for alleles of two loci, Suvar(2)1 andSuvar(3)3 (Szabad et al., 1988). In both instances mitoticrecombination in the germ-line tissue indicates that both genes arerequired for the development of the germ-line as well as the soma.Suvar(2)1 is one of the first described Su(var)s and has beenthe subject of a great deal of study. The Suvar(2)1 °' alleleincreases the transcriptional capacity of chromosomes in vitro(Khesin and Bashkirov, 1979) indicating that the gene has an effecton euchromatin function. Larvae which are heterozygous orhomozygous for the Suvar(2)1 gene exhibit hyperacetylation ofhistone H4 which is correlated with an increased accessibility ofthe DNA to endogenous endonucleases (Dorn et al., 1986). It hasbeen proposed that the gene may encode a histone deacetylase or aproduct which can control the accessibility of histones todeacetylases (Dorn et a/., 1986). In this regard, it isinteresting that strains carrying mutant alleles of Suvar(2)1exhibit reduced viability when grown on sodium butyrate (Reuter et23al., 1986; Sinclair et al., 1991), a known inhibitor of histonedeacetylase (Candido et al., 1978), while strains carrying otherstrong Su(var) loci such as Suvar(3)1, Suvar(3)2 and Suvar(3)9 donot exhibit the same sensitivity.There have been several studies describing small collectionsof enhancers of position effect variegation (E(var)s; Tartof eta/., 1984, Sinclair et al., 1989). E(var) mutations are dominantand they enhance in a general manner, their effect beingindependent of the heterochromatic breakpoint and euchromatic geneinvolved. These genes are interesting in that they may identifyproteins involved in the establishment of euchromatin domains.Cloned Su(var) loci Three dominant Su(var) loci have been cloned, Suvar(3)7,Suvar(3)6 and Suvar(2)5. Suvar(3)7 is a modifier of the haplo-suppressor/triplo-enhancer class located at 87E on the Drosophilapolytene map. This locus exhibits a dosage dependent effect fromone to five copies of the gene as determined by germ-linetransformations (Reuter et al., 1990). The encoded protein isphosphorylated, chromatin associated (G. Reuter, personalcommunication) and contains 5 widely spaced zinc finger motifswhich may allow it to interact with widely spaced segments of DNA(Reuter et al., 1990).The Suvar(3)6 locus at 87B is a modifier of the haplo-suppressor class. The locus exhibits no triplo effects even whentwo copies are introduced by germ-line transformations (G. Reuter,24personal communication).^Suvar(3)6 encodes the proteinphosphatase, PP1. Interestingly, mutations for Suvar(3)6 areepistatic over the enhancer effect of an extra copy of Suvar(3)7,a protein which contains over 40 potential phosphorylation sites(G. Reuter, personal communication).The Su(var)205 locus located at 29A is another modifier of thehaplo-suppressor/triplo-enhancer class. Cytological interval 29Ais also the location of the gene which encodes the heterochromatinassociated protein, HP1. Since two alleles of Su(var)205 havelesions in the HP1 gene, it is likely that these two genes areallelic (Eissenberg et al., 1990; G. Reuter, personalcommunication). HP1 associates specifically with regions of 0-heterochromatin although the protein sequence does not contain anyknown DNA binding motifs. A 37 amino acid region in the aminoterminus is homologous to a similar domain in the Polycomb locus(Paro and Hogness, 1991). Polycomb is a member of a family ofgenes which act as repressors of many homeotic loci including thosein the bithorax complex (Zink and Paro, 1989, Paro and Hogness,1989). Like HP1, the Pc product is a known component of chromatin(Zink and Paro, 1989) and may be part of a heterochromatin likecomplex (Paro, 1990). Interestingly, a number of the members ofthe Polycomb-like group of genes are enhancers and suppressors ofPEV (D. Sinclair, N. Clegg, T. Grigliatti and H. Brock personalcommunication).25Recessive Modifiers of position-effect variegation Although virtually all modifiers of PEV which have beendescribed exhibit dominant phenotypes, this is a reflection of theassays used, not a general characteristic of all modifying loci.There have been no screens described for recessive modifiers of PEValthough many genes (particularly those defining enzymaticfunctions) involved in determining chromatin structure would not beexpected to exhibit dominant phenotypes.Two recessive su(var) loci have been reported. The genesmfs(2)31 and wdl were identified as essential genes in division 31of the Drosophila polytene chromosome map (Sandler, 1977). Becauseof their proximity to a cluster of Su(var) genes and a secondcluster of genes which are thought to interact with heterochromatin(da-abo group; see Sandler, 1977), these mutations have been testedfor their ability to modify PEV (Sinclair et al., 1991). Althoughneither gene exhibits a dominant phenotype, both show strongrecessive suppression of white variegation in the rearrangement/n(/)W 4 .In addition to being a modifier of PEV, the mfs(2)31 strainexhibits a variety of recessive phenotypes including male andfemale sterility, abnormal bristles and prolonged development. Thefailure of flies to survive when the mutation is uncovered by adeficiency indicates that the gene also has an essential function.A number of abnormalities have been observed in the testes ofhomozygote males including, absence of motile sperm, spermatidswith two associated centrioles and spermatids with abnormally sized26nuclei (Lindsley et al., 1980). It has been proposed that mfs(2)31encodes a function required for proper centriole migration sinceimproper positioning of centrioles and spindle poles may result inmacro- and micronuclei being formed (Lindsley et al., 1980).D. SUMMARYIt is becoming increasingly apparent that an importantprinciple governing the regulation of gene expression throughdevelopment involves local alterations in chromatin structure(Gyurkovics et al., 1990). This realization has intensified thesearch for proteins which initiate and maintain such structuralchanges. Since many of these proteins will have close and strongassociations with chromatin, they may be identified by assays whichare specific for non-histone chromosomal proteins. This thesisdescribes the use of one such assay, PEV, to identify a gene whichmay encode a non-histone chromosomal protein in Drosophilamelanogaster. This protein may be involved in establishing twocytologically differentiable states of chromatin, heterochromatinand euchromatin. The mfs(2)31 gene differs from previously clonedmodifiers of PEV in that mutations exhibit no dominant phenotypes.Chapter 2 of this thesis describes the cytogenetic localization ofthe mfs(2)31 gene and the isolation of new alleles. Chapter 3contains a complete phenotypic description of the locus with aparticular emphasis on the ability of mutant alleles to modifyvariegating backgrounds. Chapter 4 contains a molecular analysisof the locus in which the putative mfs(2)31 protein sequence is27compared to previously described NHPs.28CHAPTER 2: CYTOGENETIC ANALYSIS OF SUBDIVISIONS 31D-E IntroductionThe mfs(2)31 locus is a recessive suppressor of position-effect variegation (su(var)) located in subdivision 31E of theDrosophila polytene chromosome map. Sandler (1977) placed mfs(2)31in 31B-31F, a 35 band region, based on its failure to complementDf(2L)J27 while Sinclair et al. (1991) further localized it to the20 chromomeres comprising 31D-E by its failure to complementDf(2L)J77 and Df(2L)J106.We have had an ongoing interest in the mfs(2)31 locus becauseof its su(var) phenotype. Several Su(var) genes are known toencode non-histone chromosomal proteins. Although mfs(2)31 is thefirst reported example of a recessive suppressor of PEV, theability to expand the study of this locus was hampered by severalconsiderations. First, the existence of only a single alleleseverely limits the phenotypic and developmental analysis. Second,the relative paucity of well characterized deficiencies andcomplementation groups in the 31D-E region makes it difficult toassess various strategies for a molecular analysis of the locus.To deal with these substantial limitations, three initialobjectives were identified: 1) Further cytological subdivision of31D-E by the characterization of new deficiencies to map mfs(2)31more precisely: 2) A more precise genetic, as well as cytogenetic,localization of the mfs(2)31 locus relative to previously clonedgenes in the region: 3) Isolation of new alleles of mfs(2)31 which29should include point mutations for genetic and developmentalanalysis as well as insertional mutations to aid in the molecularanalysis. It was anticipated that all these objectives could beachieved through a comprehensive cytogenetic and genetic analysisof region 31D-E; the region defined by the overlap of Df(2L)J27,Df(2L)J77 and Df(2L)J106, the smallest deficiency subinterval knownto contain mfs(2)31.The 31D-E region is contained within division 31, an anomaloussegment of the polytene chromosome colloquially referred to as the"gooseneck" because of its thin underreplicated appearance(Lefevre, 1976). Although the distinct morphology of division 31makes it a useful cytological landmark for chromosome arm 2L, theregion is poorly banded making the precise localization ofdeficiency breakpoints extremely difficult. Diffuse bandingcoupled with apparent underreplication give the region aheterochromatin-like appearance. Interestingly, division 31 hasbeen identified as a binding site for an antibody directed towardsHP1, a non-histone chromosomal protein which is primarilyassociated with heterochromatin (James et al., 1989).HP1 was first identified in a nuclear protein fraction ofDrosophila melanogaster ( James et al., 1986). A full descriptionof the gene which encodes HP1 (Su(var)205) is provided in Chapter1. Immunofluorescent labelling of an antibody specific for HP1revealed that on salivary gland polytene chromosomes, the proteinis associated with 0-heterochromatin and discrete segments ofeuchromatin including division 31 (James et al., 1989). Although30the HP1 antibody binds to division 31 with a characteristic bandingpattern, the protein contains no previously described DNA bindingmotifs. This banding could reflect the presence of repetitiveheterochromatic DNA sequences within division 31 or alternatively,HP1 could be involved in the coordinate regulation of several non-contiguous genes within this interval. These two hypotheses arenot mutually exclusive, the latter being particularly intriguingsince amongst the genes which have been localized to division 31are several known modifiers of PEV, a phenotype shared by the geneencoding HP1 (Su(var)205; Sinclair et al., 1983).There have been four genes mapped to 31D-E: daughterless (da),cdc2, 1(2)54 and mfs(2)31. The da gene was cloned and mapped by insitu hybridization to 31E (Cronmiller et al., 1988) during thecourse of the investigation reported here. It is a member of anetwork of genes which have been identified in the Drosophila sexdetermination pathway (Cline, 1989). This essential locus wasnamed for the inability of hypomorphic alleles to produce femaleoffspring. The gene has been shown genetically to be a regulatorof Sex-lethal, the primary sex-determining switch in Drosophila, aswell as an essential gene in the development of the peripheralnervous system (Cline, 1989). A molecular analysis of the locushas revealed that the da transcript encodes a transcription factorof the helix-loop-helix class (Cronmiller et al., 1988; Caudy etal., 1988). da alleles show neither dominant nor recessivesuppression of PEV and have no known functional relationship tomfs(2)31 (Sinclair et al., 1991).31Subsequent to the initiation of these studies, the Drosophilahomolog of the Yeast cdc2 gene was localized to subdivision 31E(Lehner and O'Farrell, 1990), mapping within 5 kb of the da locus(Cronmiller et al., 1988). The cdc2 family of genes encode kinaseswhose activity is required during the cell cycle both to initiateDNA synthesis (G 1-S transition) and to initiate mitosis (G 2 -Mtransition), two functions which are genetically separable in Yeast(for review see Nurse, 1990).Single alleles of 1(2)54 and mfs(2)31 were isolated in ascreen for essential genes uncovered by Df(2L)J39 (Sandler, 1977).Sandler and his colleagues were hoping to add to a previouslyidentified group of closely-linked genes which interact with Xchromosome heterochromatin. The group extends from 31D to 32D andincludes da , abnormal -oocyte (abo) and hold-up (hup). The precisenumber and distribution of genes located within this cluster cannotbe determined since no systematic genetic characterization of thislarge cytological interval has been described. 1(2)54 is arecessive lethal mutation and has not been subject to any furthergenetic analysis.As a homozygote, the mfs(2)31 mutation is semi-lethal butsterile in both sexes. Because of the recessive sterility,homozygotes cannot be tested for abnormal sex ratios amongst theirprogeny. An abnormal sex ratio amongst progeny is a characteristicof the da-abo cluster mutations. A complete description of thispleiotropic locus can be found in Chapter 3.In a subsequent analysis of a number of dominant suppressors32of PEV which map to division 31 (Sinclair et al., 1991), allmembers of the da-abo cluster were tested for an ability to modifyPEV. The genes were tested because of their apparent sensitivityto cellular levels of heterochromatin, a property shared by Su(var)genes (see Chapter 1). Although no members of the da-abo group aredominant modifiers of PEV, mfs(2)31 was found to be a strongrecessive suppressor of white variegation in In(1)wm4. Thispleiotropic effect is attributable to the mfs(2)31 locus since raresurvivors over Df(2L)J106 show the same phenotype.The mfs(2)31 locus is the first reported example of arecessive suppressor of PEV. The novelty of this gene coupled withan ability to localize it to a cytological interval (31D-E) forwhich molecular entry points (da and cdc2) are available make it anexcellent candidate for further genetic and molecular analysis.Such an analysis would be greatly facilitated by a cytogeneticstudy on the chromosomal segment in which mfs(2)31 is located. Thelocation of the gene in a region of the chromosome which, initself, is of interest to the study of Su(var)s (region of HP1binding) provides further impetus for such a study. In addition,it may be possible to determine whether the heterochromatin-likeappearance of division 31 at the cytological level correlates witha lack of mutable loci.This chapter contains a complete cytogenetic analysis of 31D-E. New deficiencies are described which divide the region intosmaller subintervals. Saturation mutagenesis brings the totalnumber of complementation groups within the region to 25, 9 of33which map to an interval containing mfs(2)31, da and cdc2. Therelative orientation of these three genes was determined withrespect to the centromere. Three new alleles of mfs(2)31 weregenerated, one of which was recovered from a screen which utilizedP element mediated hybrid-dysgenesis. As anticipated, thisanalysis made possible the further genetic and molecular analysesof the mfs(2)31 locus described in chapters 3 and 4.34Materials and MethodsStocks Unless described below, information pertaining to stocks canbe found in Lindsley and Zimm (1992). The study utilized threeseries of deficiency chromosomes generated by the reversion of theneomorphic, rumple-winged phenotype of Jammed (J). The firstseries, consisting of Df(2L)J2, Df(2L)J27, and Df(2L)J39, waskindly provided by L. Sandler (see Mange and Sandler, 1973).Series two, consisting of Df(2L)J77 and Df(2L)J106, was provided byJ. Lengyel (see Salas and Lengyel, 1984). The third series whichincluded Df(2L)J1, Df(2L)J3, Df(2L)J16 and Df(2L)J17, was generatedby D. Sinclair (unpublished). A single deficiency, Df(2L)G2, wasgenerated by the author (unpublished) in a screen for lethalalleles of Su(var)216. The list of deficiencies and theirbreakpoints, as estimated from cytological examination, are shownin Table 1.A number of female sterile mutations map to division 31 basedon their failure to complement Df(2L)J2 (Schupbach and Wieschaus,1986; 1989; 1991). The following mutant strains were kindlyprovided by T. Schupbach: trunk (trk), 1toDG25, ltoRU26, 1toPI23,erratic (err), mat(2)earlyQM47 and mat(2)synPJ50. The followingrecessive lethal and female sterile mutations were provided by L.Sandler: da, dal, mfs(2)31, hup and 1(2)54 (see Mange and Sandler,1973; Sandler, 1977). The dal allele was provided by C. Cronmillerwhile female-sterile-2-rosy-4 (fs(2)ry4) was provided by A.Spradling. The Su(var)216 mutation is described in Sinclair et35TABLE 1Cytological limits and origins of chromosomal rearrangementsRearrangement^Reference^CytologyDf(2L)J2^MANGE and SANDLER 1973^31A3-32ADf(2L)J39^MANGE and SANDLER 1973^31D-32BDf(2L)J27^SANDLER 1984^31D-31EDf(2L)J77^SALAS and LENGYEL 1984^31D-31EDf(2L)J106^SALAS and LENGYEL 1984^31D-31EDf(2L)J1 THIS STUDY^31B-31DDf(2L)J3^THIS STUDY 31D-31FDf(2L)J16^THIS STUDY^In(2L)30C-D;31Eassociated witha deficiency in31EDf (2L)J17^THIS STUDY^not visibleDf(2L)G2 THIS STUDY 31D;31F36al. (1991). A summary of these mutations and their origins can befound in Table 2.Several collections of recessive lethal mutations in the 31region have also been included in this study. The first collectionconsists of sixteen gamma-induced lethal alleles, uncovered byDf(2L)J2 (cited in Brock, 1989). Brock (1989), used ethyl-methanesulfonate (EMS) to induce a second collection of 40 lethalmutations which are also uncovered by Df(2L)J2. In a screen of5000 second chromosomes exposed to gamma-irradiation, Harrington(1990) identified 8 lethal mutations which fail to complementDf(2L)J27. One lethal allele, uncovered by both Df(2L)J2 andDf(2L)J77 was isolated by Clegg (1991) in a screen which used Pelement mediated hybrid dysgenesis as a mutagen source.Culture conditions Fly cultures were raised on a cornmeal-agarose-sucrose mediumsupplemented with tegosept as a mold inhibitor. Tetracycline andampicillin or tetracycline and streptomycin were added asantibiotics. All experiments were performed at 25° unlessotherwise indicated.Cytology Black cells (Bc) is a dominant larval marker on the secondchromosome. Males carrying a Bc chromosome and the chromosome ofinterest, were crossed to wild-type virgin females at 17°. Larvaecarrying the mutant chromosome (Bc*) were selected for cytological37TABLE 2Origin of previously described mutations in region 31A-32AMutation^ ReferenceSu(var)216^SINCLAIR et al., 1983; 1991tip^DE VALOIR et al., 1991trk SCHUPBACH and WIESCHAUS, 1986bsk^NUSSLEIN-VOLHARD, WIESCHAUS and KLUDING, 1984itoRU26^SCHUPBACH and WIESCHAUS, 1991ltoDG25 SCHUPBACH and WIESCHAUS, 1991da^ CLINE, 1989mfs(2)31^LINDSLEY, GOLDSTEIN and SANDLER, 1980pim NOSSLEIN-VOLHARD, WIESCHAUS and KLUDING, 19841(2)54/^SANDLER, 1977; SCHUPBACH and WIESCHAUS, 1989mat (2)earlyQM4 7err^SCHUPBACH and WIESCHAUS, 1989mat(2) SCHUPBACH and WIESCHAUS, 1989synPJ501toPI23^SCHUPBACH and WIESCHAUS, 1991cdc2 SINCLAIR et al., 1991wdl^SANDLER, 1977hup SANDLER, 1977dal^SANDLER, 197738examination. Salivary glands were dissected from 3rd instar larvaein Drosophila ringers (6.5 g NaC1, 0.14 g KC1, 0.12 g CaC1 2 , 0.20g NaHCO 3 , 0.01 g NaH 2PO 4 in 1000 ml dH 2O; pH 7.2) and were thentransferred immediately to a drop of aceto-alcohol (3 partsabsolute alcohol: 1 part glacial acetic acid) on a clean acid-washed slide. Following a two minute fixation in the aceto-alcohol, a drop of aceto-orcein stain (3% in 45% acetic acid) wasadded directly to the glands. Glands were stained for 20 minutesand then excess stain was removed by rinsing carefully with 45%acetic acid. A single drop of lacto-aceto-orcein stain (3% in 1:1glacial acetic acid and lactic acid) was added followed by acoverslip. The glands were squashed and then examined under phasecontrast optics.Genetic screens Three genetic screens were conducted to isolate new lethalmutations in the 31D-E region. The protocols for these screens aresummarized in Figures 1 and 2.The first two screens were designed to generate EMS inducedlethal mutations uncovered by Df(2L)J27 and Df(2L)J106 respectively(see Figure 1). Males which were homozygous for a geneticallymarked second chromosome (b pr cn), were fed EMS (0.025 M) by themethod of Lewis and Bacher (1986), and mated, en masse, to Tft/CyOfemales. Male progeny with the genotype b pr cn/CyO were collectedand individually mated (at 29°) to harems of 3-5 Df(2L)J27/Cy0(screen #1) or Df(2L)J106/CyO (screen #2) virgin females. The39Figure 1: Screens for EMS induced lethal mutationsMales^ FemalesEMS25 o^b pr cnb pr cn^ CyODf(2L)J2729 o^b pr cn^X^or CyOX^TftScreen 1CyO Df(2L)J1 06 Screen 2 CyO*b pr cnDf(2L)J27^Df(2L)J27or orDf(2L)J1 06^Df(2L)J 1 06-CyO^CyO b pr cnEstablish^ Score forBalanced Absence.Stocks40Males^ Femalesb pr cn^+ ^, ^Sp^.X ^b pr cn +^CyO)CyO^+^birmAb pr cn^LV-3 SbX ^Tftbirm Of^CyOA*b pr cn^+^Df(2L)J77X ^ ^, ^Figure 2: Screen for^P element induced lethal mutations46.2-3 Sb 25 °TM6 Ubx16 °Of+16 °++ ^2 5 °CyO Of CyO^+^/^ I* *b pr cn ^Df(2L)J77^b pr cn ^, ^^CyO^CyO^Df(2L)J77Establish Score forBalanced^ Absence.Stocksb pr cn^. L■2-3 SbX ^birm^Of, ^ ^41presence of a lethal mutation on the b pr cn chromosome, which isuncovered by either of the two deficiencies, was indicated by theabsence of b pr cn/Df(2L)J27 or Df(2L)J106 progeny (Cy). Stocksof putative lethal mutations were established from balancedsiblings (b pr cn/CyO) and were re-tested at 22° to identifytemperature-sensitive mutations.A third screen was conducted using P element mediated hybriddysgenesis to generate insertional lethal mutations uncovered byDf(2L)J77 (Figure 2). This screen utilized the A2-3 mutagenizingsystem as described by Robertson et al., (1988). Males which werehomozygous for the b pr cn chromosome were mated en masse tofemales with the genotype Sp/CyO; 02-3 Sb/TM6, Ubx. The gene 02-3is a P factor with the inverted repeats deleted; it synthesizestransposase but cannot move. The b pr cn/+; A2-3 Sb/+ males werecollected and mated en masse to birm/birm; ry/ry females (the birmstrain contains approximately 17 P elements which can respond totransposase when it is provided from another source). This crossand the next were done at 16° to minimize somatic mobilization ofP elements. Dysgenic males of genotype b pr cn/birm; A2-3 Sb/rywere collected for mass matings to Tft/CyO females. The b prcn/CyO; ry/+ males were then collected and individually mated toharems of Df(2L)J77/Cy0 virgin females (25'). As was the case inscreens #1 and #2, a lethal mutation was indicated by the absenceof b pr cn/Df(2L)J77 progeny. Stocks of putative lethal mutationswere established from balanced siblings.42Complementation New and previously identified mutations in division 31 werecrossed to the collection of deficiencies available for the region.Alleles were assigned to a cytological interval based on eitherlethality, sterility or a visible phenotype. Failure to complementwas based on a minimum of 50 flies scored although in the vastmajority of cases 200-300 flies were examined. Once placed withina cytological interval, alleles were sorted into complementationgroups by inter se crosses. If a cytological interval wasdetermined by earlier rounds of crosses to have a large number ofcomplementation groups, a strong representative allele from eachgroup was designated as the tester strain for subsequent crosses.In the case of some sterile mutations, assignment to a group wasbased on a failure to complement a single allele.Mapping by Recombination In order to determine the order of complementation groups withrespect to the centromere, lethal strains carrying differentgenetic markers on the mutant chromosome were crossed. Femalesheterozygous for both mutations were collected and crossed, enmasse, to either Df(2L)G2/Cy0 or Df(2L)J77/Cy0 males. Putative Cy'recombinants were identified by the presence of straight-wingedflies. These were collected and retested for their ability tocomplement the deficiency chromosome. Gene order was inferred bythe genetic markers on the recombinant chromosome. The geneticmarkers used were black (b) , purple (pr) and cinnabar (cn) located43at map positions 48.5, 54.5 and 57.5 respectively (Lindsley andZimm, 1992), all of which are proximal to the 31 region. Theclosest marker, b, is located at map position 48.5. In some casesonly two recombinant chromosomes were recovered.44ResultsCytology Five new deficiencies within division 31 were identified andcharacterized: Df(2L)J1, Df(2L)J3, Df(2L)J16, Df(2L)J17, andDf(2L)G2. The breakpoints for these deficiencies are listed inTable 1 along with the previously described deficiencies for theregion: Df(2L)J2, Df(2L)J39, Df(2L)J27, Df(2L)J77 and Df(2L)J106.As previously mentioned, division 31 has a very diffuse bandingpattern making the precise determination of rearrangementbreakpoints very difficult. Nonetheless, careful cytologicalexaminations coupled with recent molecular localizations of geneswithin this cytological division have facilitated the drawing of anew, reasonably accurate, cytological map (Figure 3). Non-visibledeficiencies (i.e. Df(2L)J17) have not been placed on this mapalthough their breakpoints can be inferred from strictly geneticconsiderations. What follows is a rationalization of thebreakpoints which have been determined for the new deficiencies.Df(2L)J1 is a visible deficiency contained completely withindivision 31. Cytologically, it removes the distal two thirds of31D which sets the proximal breakpoint in proximal 31D. This isconfirmed genetically by its failure to uncover the da gene, whichhas been localized by in situ hybridization to distal 31E(Cronmiller et al., 1988). Although the distal breakpoint fallswithin the extremely poorly banded 31B-C region, the failure of thedeficiency to complement the fs(2)ry4 gene, which was localized byin situ hybridization to 31B (personal observation), makes 31B the45Figure 3: Cytological map for division 31. Dashed lines represent regions of uncertainty.Df(2L)J2Df(2L)J39Df(2L)J77Df(2L)J106Df(2L)J27Df(2L)J1Df(2L)J3Df(2L)J16Df(2L)G2 most likely site of the distal breakpoint.Df(2L)J3 is a visible deficiency in the proximal end ofdivision 31. Proximally, it deletes the sharply banded 31F1 butnot the 32A1 doublet thus placing the proximal breakpoint within31F. The distal breakpoint does not appear to remove the proximalmost band of 31D although genetically, the deficiency uncovers theda gene. This suggests that the distal breakpoint is in 31E butvery close to the 31D-E boundary.Df(2L)J16 is a small deficiency within subdivision 31E. It isassociated with an inversion whose distal breakpoint is at the 30C-D boundary and whose proximal breakpoint falls within 31E. Thethree distinct bands which comprise all of 31D are clearly visiblewithin the inversion loop thus setting the distal deficiencybreakpoint near the 31D-E boundary. This was confirmed geneticallyby the failure of the deficiency to complement the da gene. The31F1 band is not deleted thus localizing the deficiency to 31E.Df(2L)J17 is not visible. Both the proximal most band of 31Dand band 31F1 are not removed by this deficiency. Genetically, itappears to be smaller than Df(2L)J16, uncovering the distal mostcomplementation groups.Df(2L)G2 is the only known deficiency for the 31 region whichwas not isolated by reverting the dominant phenotype of theneomorphic mutation Jammed (J). It was isolated in a screen forgamma induced lethal alleles of Su(var)216 (unpublished). It is asmall visible deficiency in the proximal end of the 31 region. Itsdistal breakpoint falls within 31D, although a more precise47localization was not possible. Genetically, it appears to have avery similar distal breakpoint to Df(2L)J39, uncovering all of thesame complementation groups. Proximally, it deletes 31F1 but notthe 32A doublet placing the proximal breakpoint within 31F.Screens Lethal mutations in the 31 region were generated in sevenindependent screens conducted in our lab. Four of the screens havebeen described previously while three were performed as part ofthis current study. The results of the new screens are summarizedin Table 3.Screen #1 used the potent mutagen ethyl methanesulfonate (EMS)to generate lethal mutations uncovered by Df(2L)J27. A total of10,000 chromosomes were examined from which 27 lethal alleles wererecovered (frequency = 0.0027). Screen #2 used EMS to generatelethal mutations uncovered by Df(2L)J106. This screen alsoproduced 27 mutations but in only 5000 chromosomes examined(frequency = 0.0054). None of the mutations recovered from Screens#1 and #2 are temperature-sensitive.Screen #3 was designed to generate insertional mutations inthe 31D-E region. The A2-3 mutagenizing system (Robertson et al.,1988) was employed to jump P elements into the b pr cn chromosome.Twelve thousand five hundred chromosomes were screened for failureto complement Df(2L)J77. Four lethal mutations were recovered fora mutation frequency of 0.0003.48TABLE 3Summary of genetic screensScreen TesterstrainMutagen ChromosomesscreenedMutationsrecoveredMutationFrequency1 Df(2L)J27 EMSa 10000 27 0.00272 Df(2L)J106 EMS 5000 27 0.00543 Df(2L)J77 HDb 12500 4 0.00032Total 27500 58a ethyl methanesulfonateb P element hybrid dysgenesis49Complementation The 58 new alleles were combined with 65 previously identifiedlethal mutations in division 31 (see Materials and Methods) and 19mutations isolated on the basis of other phenotypes (see Materialsand Methods and Table 2), to generate a pool of 142 testablestrains. Mutations were initially screened for those which failedto complement Df(2L)J106, Df(2L)J77 and Df(2L)J27. The overlapbetween these three deficiencies defines the smallest cytologicalinterval to which the mfs(2)31 gene had been previously assigned.For the purposes of this study, only mutations which failed tocomplement all three deficiencies were examined further. A totalof 97 mutant strains fell into this category. These were thentested against the newly characterized deficiencies in the region:Df(2L)J16, Df(2L)J1, Df(2L)J17, Df(2L)J3 and Df(2L)G2. Three ofthese deficiencies (Df(2L)J16, Df(2L)J17 and Df(2L)J1) appear,genetically, to have breakpoints within the overlap of Df(2L)J106,Df(2L)J77 and Df(2L)J27 (see Figure 4). Thus, it was possible toassign all 97 mutant strains to one of six cytogenetic intervals(intervals 1-6 in Figure 4).The mutations were broken down by inter se complementationinto 25 complementation groups, 9 of which are represented by asingle allele. The smallest intervals, #1 and #4, contain singlecomplementation groups while the largest interval, #5, contains 9.Fourteen of the complementation groups have not been previouslyidentified. New alleles were generated for ltoDG25, err, pim, da,cdc2, 1(2)54 and mfs(2)31. ltoDG25 and err had been previously50Figure 4: Cytogenetic map of region 31 D-E. Lethal complementation groups have been named byconventional nomenclature but are abbreviated in the figure i.e.^1(2)31Db = Db .Total number of alleles and commonly used synonyms are provided in parentheses.31D^ 31E^ 31FDb(4)1 2RU26DcDd4bsk^Of^Ea^Eb^Ej^Ek^I (Pim; 5) (da;10) (10)^(3) (1(2)54;DG25 I 8)(8)^Ec^Ed^PJ50^El(5)^(4) (err;4)3 5^6De Ee^Ef^P123^Em(4)^(3) (2)Eg^Eh^En^Eo(Cdc2;^(2)12)Ei(mfs(2)31;4)Df(2L)J77Df(2L)J27 Df(2L)J 106Df(2L)J17Df(2L)J 16Df(2L)J 1identified as maternal effect female sterile mutations.^Theisolation of lethal alleles of these genes indicates an essentialfunction not restricted to the female germline.The mfs(2)31 gene was localized to the largest subinterval:#5. This interval contains nine complementation groups includingthe previously cloned genes, da and cdc2. All groups within thisinterval exhibited a simple complementation pattern. There are twomutations which were generated in P element screens which map tothis interval. The single mutation generated by Clegg (1991) is inthe cdc2 complementation group while a second mutation isolated inscreen #3 (this study) was assigned to the mfs(2)31 complementationgroup. In total, three new alleles of mfs(2)31 were generated, onein a P element screen (screen #3) and two in an EMS screen (screen#1). These new alleles are described in detail in Chapter 3.The interval to which mfs(2)31 was localized is defined by theproximal breakpoints of Df(2L)J16 and Df(2L)J17. The next mostproximal interval is defined by the proximal breakpoints ofDf(2L)J27 and Df(2L)J16. This interval, with its eightcomplementation groups, is contained completely within subdivision31E (see Figure 3). If a random dispersal of genes is assumedwithin 31E, the data indicates that mfs(2)31 is located in thedistal half of this subdivision. This is supported by theobservation that the previously cloned genes cdc2 and da, both inthe same interval as mfs(2)31, have been localized by in situhybridization to 31E1-E3 (Cronmiller et al., 1988; Lehner andO'Farrell, 1990).52Interval 31D-E can be defined approximately by the overlapbetween Df(2L)J27, Df(2L)J106 and Df(2L)J77. All screens describedwithin this study overlap within this region. However, because allscreens were done independently using a variety of deficiencies andmutagens, an empirical determination of the degree of saturation ofthis region is not possible. Nonetheless, the existence of 9complementation groups out of 25 with only one allele would suggestthat the region is not saturated. This is supported by the recentidentification of new complementation groups within the region(Randy Mottus, personal communication).Mapping by Recombination The smallest cytological interval into which mfs(2)31 can beplaced (interval #5 on Figure 4) contains two previously clonedgenes: da and cdc2. The position of these two genes relative toone another has been determined molecularly (Clegg, 1992), but notgenetically. Genetically, mfs(2)31 has been mapped proximally toda but has not been mapped relative to cdc2 (Lindsley et al.,1980). A precise positioning of mfs(2)31 relative to the twocloned genes could be of great value for the molecular analysis ofthis gene. More specifically, it would allow for the orientationof a molecular walk.Genetic mapping was used to determine the order of da, cdc2and mfs(2)31 relative to the centromere. The results of theindividual pair-wise mapping experiments are provided in Table 4and these are summarized in the genetic map provided in Figure 5.53TABLE 4Mapping by recombinationCross FliesScoredRecomb. Map'DistanceOrderpim x cdc2 30562 7 0.046 pim - cdc2Eb^x cdc2 78755 2 0.005 Ec^- cdc2da^x cdc2 39270 2 0.010 cdc2 - damfs(2)31 x da 66181 8 0.024 da - mfs(2)31Ef x cdc2 31128 7 0.045 cdc2 - EfEf x mfs(2)31 58775 5 0.017 mfs(2)31 - Efa map distance = ((Recomb. X 2)/Parentals) X 100. The numberof parental types was estimated from the number of Cy siblings.54Figure 5: Recombination map for complementation groups located in cytogenetic interval#5. More commonly used synonyms are provided in parentheses. pim is locatedin cytogenetic interval #4.0.005^0.01^0.024^0.017-^ - - -^ - - - -1(2)31 Df^1(2)31 Ec 1(2)31 Eh^1(2)31 Ea^1(2)31 Ei^1(2)31 Ef(pim) (cdc2) (da) (mfs(2)31 )0.046^-^ CentromereAn initial mapping experiment was done with pim and cdc2, two geneslocated within adjacent cytological subintervals. This was done toverify the orientation of the cytogenetic map relative to thecentromere. Although this orientation can be inferred by thepattern of complementation with respect to Df(2L)J1, thedifficulties with cytology in the region made this verificationadvisable. As predicted by the cytological map, pim maps distal tocdc2 (0.046 map units). The mfs(2)31 mutation has previously beenmapped proximal to da, although no map distance has been reported(Lindsley et al., 1980). We confirm this orientation placingmfs(2)31 0.024 map units proximal to da. A mapping experimentbetween da and cdc2 yielded only two recombinants placing cdc2 0.01map units distal to da. This low frequency of recombinationreflects the physical distance between these two genes < 3 kb.This order was confirmed recently in an unrelated study in whichthree recombinants were recovered (D. Sinclair, personalcommunication).In an unrelated study, I am conducting a complete ordering ofthe complementation groups in subdivision 31E. At the current timetwo additional complementation groups have been placed on the map:1(2)31Eb and 1(2)31Ef. Their map positions relative to da, cdc2and mfs(2)31 are indicated in Figure 5.56DiscussionTo aid in the genetic and molecular characterization of themfs(2)31 locus, a cytogenetic analysis has been conducted in thesmallest cytological interval to which it has been localized, 31D-E. This adds 31D-E to a growing number of regions in theDrosophila genome which have been intensively characterized in thismanner. There are three objectives which can be met through suchan analysis, a more precise cytological localization of the gene,an orientation of the locus with respect to molecular markers inthe region and the isolation of new mutant alleles, particularlyones which are the result of insertional events. In addition, acytogenetic analysis within division 31 may provide a clue as towhy the region has specific associations with HP1, aheterochromatin associated protein.The majority of 31D-E is encompassed by the overlap ofDf(2L)J27, Df(2L)J106 and Df(2L)J77. The elaboration of acytogenetic map for this region involved two steps, the cytologicalcharacterization of new deficiencies for the region and theplacement of new and previously identified loci onto an updatedcytological map. It was hoped that new deficiencies would sub-divide the relatively large cytological interval into whichmfs(2)31 had been mapped. From amongst a collection of revertantsof the Jammed phenotype several new rearrangements were identifiedwith breakpoints in the 31D-E region . These new rearrangementswere combined with previously described ones to generate a newcytological map for division 31 (Figure 3 and Table 1).57A total of 142 mutations were examined for complementationwith the battery of rearrangements described in Table 1. Nineteenof the mutations were provided by other laboratories (Table 2),sixty-five are from our collections and fifty-eight are newmutations induced in this study. The mutation frequencies for thenew EMS screens are 0.0027 and 0.0054. A survey of recentliterature indicates that these frequencies are typical for f 2 EMSlethal screens over similarly sized euchromatic deficiencies(0.0026, Wohlwill and Bonner, 1991; 0.009, Belote et al., 1990;0.0036, Kimble et al., 1990). A screen of 12,500 secondchromosomes exposed to P element mutagenesis identified four lethalmutations uncovered by Df(2L)J77. The A2-3 system has been shownto be a potent mutagenizing system producing 1-3 insertional eventsper chromosome arm per generation (Robertson et al., 1988). Thispredicts over 200 insertional events into 31D-E in a screen of thismagnitude. The low rate of recovery of lethal mutations cannot beattributed to a lack of genetic activity within the region sincethere are at least 28 lethal complementation groups uncovered byDf(2L)J77. A previous screen of 14,000 chromosomes utilizing Pelement mediated hybrid dysgenesis generated only two mutationalevents uncovered by Df(2L)J77 (Clegg, 1992). Although the resultsof these two screens would suggest that region 31D-E may besomewhat refractory to insertional mutagenesis, there is no goodempirical data available which would suggest that the observedmutation frequencies are atypical for the Drosophila genome.A new cytogenetic map was generated for region 31D-E (Figure584) providing valuable information for further studies on mfs(2)31.With regard to a more precise cytogenetic localization of thissu(var) locus, the 31D-E interval is now divided into six smallercytological subintervals containing as few as one and as many asnine lethal complementation groups. A total of twenty-fivecomplementation groups were identified, nine of which are definedby a single allele. This large number of complementation groupsrepresented by a single allele suggests that the region has notbeen saturated for essential genes. This was recently verified bythe isolation of new lethal complementation groups within theregion (R. Mottus, personal communication). The mfs(2)31 gene waslocalized to a genetically large subinterval containing eight othercomplementation groups. These include the previously cloned genesda and cdc2, both of which have been localized by in situhybridization to distal 31E (Cronmiller et al., 1988; Lehner andO'Farrell, 1990). Since at least 1/2 of the complementation groupswhich map to 31E are proximal to mfs(2)31, the locus is probablylocated in the distal half of this subdivision.Since mfs(2)31 is contained within the same cytogeneticinterval as two previously cloned genes, da and cdc2, theorientation of these three genes was determined with respect to thecentromere. The gene order, moving distally to proximally, is cdc2- da - mfs(2)31. The map distance between da and mfs(2)31 isroughly three times greater than the distance between da and cdc2,although the cdc2 - da distance was based on a low number ofrecombinants recovered. Since the distance between cdc2 and da is59less than 3 Kb (Clegg, 1992), we infer that mfs(2)31 is quite closeto these two genes on the molecular map. A genomic walk from da,away from cdc2, would cross the proximal breakpoint of Df(2L)J17,defining the segment of DNA within which the mfs(2)31 locus shouldbe contained.Although a number of the complementation groups contain 10 ormore alleles only three additional alleles of mfs(2)31 wereidentified. Two alleles were recovered in the new EMS screens(Figure 1) while a third allele was isolated in the P elementscreen (Figure 2). The isolation of a P induced allele of mfs(2)31was fortuitous since our analysis of the region indicates thatalthough it is very active genetically, it may be refractory to Pelement mutagenesis. These new mutations, along with Sandlers(1977) original allele, are described in detail in Chapter 3.Prior to this study, very few genetic functions had beenconfirmed in division 31. This was of interest since not only doesthe region have a cytological morphology reminiscent of 0-heterochromatin but it is also one of the few segments ofeuchromatin which is associated with HP1, a Drosophila proteinwhich associates primarily with regions of j3-heterochromatin (Jameset al., 1989). However, unlike heterochromatin, division 31 doesnot appear to have a lower level of genetic activity than othersegments of euchromatin. Twenty-five lethal complementation groupshave been identified in 31D-E alone. This number is in excess ofthe bands identified in the region although it is not atypical forDrosophila euchromatin (Lefevre and Watkins, 1985). In terms of60total gene numbers, patterns of complementation, accessibility tomutagens and frequency of recombinational events, division 31 ismore typical of euchromatin than heterochromatin. Although, thiscytogenetic analysis does not provide any indication why HP1 wouldassociate specifically with this region, the level of geneticactivity demonstrated for 31D-E may not be typical for all ofdivision 31.Based on this cytogenetic analysis, a comment can be made withregard to the relationship between mfs(2)31 and the remainder ofthe da-abo region cluster of genes (Sandler, 1977). The mfs(2)31locus has been included in this cluster based on its physicalproximity to the other genes and its phenotypic similarity toanother member of the group, hup. The results of this study bringinto question the first of these two criteria. Although, the da andmfs(2)31 genes, both members of the proposed cluster, map to thesame cytogenetic interval they are genetically separable from theremainder of the cluster by a minimum of ten complementationgroups. Although these could represent as yet undescribed membersof a functionally related group of genes, the size of such acluster would be unprecedented.61CHAPTER 3: GENETIC ANALYSIS OF THE mfs(2)31 LOCUS IntroductionPosition-effect variegation (PEV) occurs when a chromosomalrearrangement establishes a close association between a brokenpiece of heterochromatin and a euchromatic gene (for review seeChapter 1). As a consequence of this rearrangement, the gene istranscriptionally inactivated in some cells while remainingcompetent in others. The determination of transcriptionalcompetence is made early during development and, for the most part,is propagated clonally. If the affected gene encodes a productwhich acts in a cell autonomous manner, a variegated phenotype maybe observed. Although the molecular basis of the inactivatingevent is not known, the inactivation is not a consequence ofsomatic mutation since complete wild-type activity can be restoredto a variegating gene by removing it from the heterochromaticbreakpoint (Judd, 1955). The most durable hypothesis to dateproposes that inactivation results from the aberrant packaging ofthe variegating gene as heterochromatin (Prokofyeva-Belgovskaya,1941). The involvement of chromatin in the variegating phenotypeis well illustrated by the sensitivity of the phenomenon to theploidy of sex chromosome heterochromatin (Gowen and Gay, 1934) andthe histone gene cluster (Khesin and Leibovitch, 1978; Moore etal., 1979).Since its earliest description, it has been proposed that thePEV phenomenon may provide an effective genetic assay for genes62involved in the establishment and maintenance of different statesof chromosomal compaction. As is the case with the histone genes,the variegating phenotype may be sensitive to the ploidy of otherloci which encode structural components of chromatin. Severallarge collections of single-locus modifiers of PEV have beendescribed (Reuter and Wolff, 1981; Sinclair et al., 1983; Reuter etal., 1986) of which the best characterized are the dominantsuppressors of PEV, the Su(var) family. Mutant alleles of Su(var)genes, when placed in trans with a variegating rearrangement,exhibit strong, dominant suppression of PEV, restoring thephenotype to an almost wild-type appearance. It has been proposedthat at least a subset of the Su(var) genes will encode non-histonechromosomal proteins (NHPs) which are responsible for theestablishment and maintenance of heterochromatin (Spofford, 1976)-A reduction in the availability of NHPs would reduce theprobability that any given cell can produce sufficientheterochromatin to inactivate the variegating gene.This chapter contains a phenotypic analysis of a novel type ofrecessive su(var) locus, mfs(2)31. mfs(2)31 is a male and femalerecessive sterile mutation which was isolated in a screen foralleles of the maternal-effect genes daughterless (da; Bell, 1954)and abnormal-oocyte (abo; Sandler et al., 1968; Sandler, 1977).Sandler and his colleagues (1968) had observed a phenotypicsimilarity between hypomorphic alleles of the da and abo genes.Mothers which are homozygous for hypomorphic da alleles produce nofemale progeny while mothers which are homozygous for hypomorphic63abo alleles, and which carry the normal complement of sexchromosomes, produce primarily female progeny if crossed to malescarrying attached XY chromosomes (no free Y). Both maternaleffects can be modified by varying the amount of sex chromosomeheterochromatin in either the mutant mother or her offspring.Sandler proposed that both of these loci modify the function ofgenes located within heterochromatin either by regulating theirexpression or by interacting with their products directly (Sandler,1968).In order to isolate new alleles of da and abo, Sandler (1977)screened for EMS-induced lethal mutations uncovered by Df(2L)J39.Both da and abo mutations fail to complement this large deficiencyfor divisions 31 and 32 of chromosome arm 2L. In a screen of 455second chromosomes, single alleles of five new loci were isolatedand described: da-abo like (dal), wavoid-like (wdl), hold-up (hup),1(2)54 and mfs(2)31 (originally designated mfs48). In the hup,wdl, and dal strains the progeny of homozygote mothers exhibit anabnormal sex ratio, a phenotype which is sensitive to the maternalX heterochromatin constitution. Because of the similar phenotypesand the close-linkage of hup, wdl, dal, da and abo, Sandler (1977)proposed that they constituted a large cluster of functionallyrelated genes within the so-called da-abo region on chromosome arm2L.Because of their respective homozygous lethal and homozygoussterile phenotypes, the 1(2)54 and mfs(2)31 strains could not betested for abnormal sex ratios or sensitivity to maternal X64heterochromatin constitution. However, in a subsequent study(Lindsley et al., 1980), mfs(2)31 was added to the da-abo clusterbased on two criteria. First, recombination mapping placesmfs(2)31 between the da and abo genes, thus physically localizingit to the cluster. Second, a more detailed analysis of the mutantstrain reveals some phenotypic similarities with hup, a previouslydescribed member of the group. Both hup and mfs(2)31 are lethal ashemizygotes but semi-lethal as homozygotes. Viable homozygotes arecompletely sterile at 28.5° while showing varying degrees offertility at 23°. The testes of male homozygotes have beenexamined in both mutant strains. At 23°, spermatogenesis appearsnormal, the testes are full of cells and the seminal vesiclescontain motile sperm. In contrast, at 28.5° there are no motilesperm in the seminal vesicles of either mutant strain. Inaddition, a variety of defects are apparent in the later stages ofspermiogenesis, beginning shortly after meiosis. Spermatids areobserved with both macronuclei and micronuclei and there areoccasional nuclei with two basal bodies attached. In wild-typestrains, spermatids contain a single nucleus of uniform size andeach spermatid receives only a single centriole which attaches tothe nucleus and acts as a basal body. Lindsley and his colleagues(1980) propose that both mutations have defects in the systemsresponsible for segregation of the centrioles during meiosis.Abnormally sized nuclei would result from an incorrect alignment ofspindle poles and centrioles prior to division. Since the mfs(2)31locus has a similar phenotype to hup and is located between da and65abo, it was included in the da-abo cluster.In an unrelated study (Sinclair et al., 1991), a group ofdominant suppressors of PEV had been mapped to division 31 and wereuncovered by Df(2L)J39. Since Su(var) genes are sensitive tointracellular levels of heterochromatin, a phenotype shared by theda-abo region cluster, it was of interest to know whether membersof the cluster exhibit a Su(var) phenotype. The mutant strainshave been tested, both as heterozygotes and as homozygotes, for anability to suppress white variegation in In(l)w". This inversionon the X chromosome moves the white gene (w) from its distaleuchromatic location to a position adjacent to a breakpoint withincentric heterochromatin. Since the result is a mottled eye, thisrearrangement serves as an excellent reporter gene for the actionof second-site modifiers of PEV. None of the da-abo region clusterof genes show any appreciable dominant suppression when placed intrans with the In(l)w" rearrangement. However, two members of thegroup, mfs(2)31 and wdl, show strong suppression as homozygotes.The phenotype in the mfs(2)31 strain is attributable to themlfs(2)31 locus since rare mfs(2)31/Df(2L)J106 survivors alsoexhibit strong suppression.mfs(2)31 and hup are the first reported examples of recessivemodifiers of PEV. To date, there has been no attempt to screen forsuch mutations, primarily because of the ease with which dominantmodifiers can be obtained. The standard assay for modifiers of PEVhas been to screen for dominant enhancers and suppressors of the te 4variegating rearrangement. Large collections of modifiers have66been isolated in this manner, and several loci have now beenexamined molecularly (reviewed in Chapter 1).In addition to the building blocks necessary to produceheterochromatin, a cell would require enzymatic functions toassemble, disassemble and maintain the integrity of the complex.Throughout the cell cycle, chromatin exhibits a variety of changesin degrees of compaction although the systems responsible foreffecting these changes have not yet been identified. There isalso mounting evidence that fundamental developmental processessuch as differentiation, determination, and maintenance oftranscriptional competence are intimately tied to alterations inchromatin environment (Pfeifer et al., 1987; Paro, 1990). It isunlikely that all the enzymatic functions required to givechromatin its dynamic character would be contained within thestructural elements themselves. Hence, some of these transientfactors in chromatin packaging may not necessarily be identifiableby dominant suppression of PEV. Nonetheless, PEV could be used asan assay for genes involved in these processes. The genes whichencode proteins which modify chromatin through more transientenzymatic interactions are likely to differ from previouslydescribed Su(var) genes in two ways. First, unless such proteinsare involved in rate-limiting steps in enzymatic pathways,mutations within these genes are more likely to exhibit recessiverather than dominant phenotypes with respect to PEV. Second,important modifiers of chromatin assembly could have such aprofound impact on cellular activity that they are likely to be67essential loci, only exhibiting a su(var) phenotype in viablehypomorphic alleles. Both criteria are met by the mfs(2)31mutation.This chapter describes mfs(2)31, a novel type of su(var) locuslocated in cytological subdivision 31E.^In addition to theoriginal mfs(2)31 mutation (mfs(2)31 1), three new alleles(mfs(2)31 2 , mfs(2)31 3 and mfs(2)31 4 ) have been isolated (see Chapter2). The four alleles will be referred to hereafter as mfs 1 , mfs 2 ,mfs 3 and mfs 4. The mfs 2 allele was induced in a screen for Pelement insertions while mfs 3 and mfs 4 were induced in screenswhich utilized EMS as a mutagen. The isolation of new alleles ofmfs(2)31 makes possible a more complete genetic description of thispleiotropic locus.Although none of the mfs(2)31 alleles showed dominantphenotypes with respect to PEV, surviving combinations of mfs(2)31hypomorphic alleles exhibited recessive suppression of PEV in allthe variegating backgrounds in which they were tested. Anexamination of ovaries from homozygous females revealed that thefemale sterility is associated with a delay in vitellogenesis.Although the gene is early acting (all alleles exhibiting larvallethality), an examination of the sterility phenotype under avariety of temperature regimes indicates that the gene product mayhave an adult function. A high frequency of reversion of mfs 2 wasobserved when the allele was outcrossed under dysgenic conditions.This allele carries a P element insertion in subdivision 31E whichhas been used for the molecular analysis of the mfs(2)31 locus68described in Chapter 4.69Materials and MethodsStocks Descriptions of all special chromosomes and mutations used inthis study can be found below and/or in Lindsley and Zimm (1992).mfs 1 is an EMS-induced male and female sterile mutation which mapsto division 31 because of its failure to complement Df(2L)J39 andDf(2L)J77. This mutant strain was kindly provided by L. Sandler.The isolation of three new alleles of mfs(2)31 is described inChapter 2: mf.52 , mfs 3 and mfs 4 . mfs 2 was recovered in a P elementscreen while mfs 3 and mfs 4 are EMS-induced. All three mutationswere induced on an isogenic b pr cn chromosome. Df(2L)J77 is avisible deficiency for region 31D-E which has been describedelsewhere in this study (see Table 1).Two collections of second chromosome Su(var) mutations havebeen tested for allelism with mfs(2)31. The first series consistsof strongly suppressing alleles which were isolated in an EMSscreen for dominant modifiers of PEV (Sinclair et al., 1983):Su(var)204, Su(var)205, Su(var)212, Su(var)213 and Su(var)214. Thesecond series are weak Su(var) mutations which have been mapped todivision 31 by recombination (G. Reuter, personal communication):Sup20, Sup34, Sup37, Sup39, Sup41, Sup46 and Sup47. The firstseries was selected from our lab collection while the second serieswas kindly provided by G. Reuter.In ^pe 51b, In(1)W' and In(1)W 4 (hereafter referred to as Wm51b ,w'T and Wm' respectively) are X chromosome inversions which variegatefor the distally located white (w) gene. In the cases of winsm and70W 4 , the variegation inducing regions remain near the centromere andthe affected euchromatic loci are brought near them while in wn' thevariegation inducing region is removed from the centromere andbrought near the distally affected loci. Additional information onthe breakpoints of these rearrangements can be found in Tartof etal. (1984). The Wm51b and wm' strains rapidly accumulate modifiersin laboratory stocks showing considerably weaker variegation thanis observed following several rounds of out-crossing. Tocircumvent this problem, chromosomes were used which had beenmaintained as heterozygotes with a w chromosome for seventeengenerations (T4r/y x Wmottled/w- ) and which had been selected at eachgeneration for strong variegation. These strains were kindlyprovided by V. Lloyd.Culture conditions Fly cultures were raised on a cornmeal-agarose-sucrose mediumsupplemented with p-aminobenzoic acid as a mold inhibitor.Tetracycline and ampicillin or tetracycline and streptomycin wereadded as antibiotics. All experiments were performed at 25° unlessotherwise indicated.Complementation Genetic complementation among strains was determined asfollows. Two mutant strains, each carrying the dominantly markedsecond chromosome balancer In(2LR)Cy0 (C1/0), were crossed. Failureto complement was indicated by the absence of Cy' progeny in a71minimum of one hundred and fifty flies scored. Viability isexpressed as Cy' flies recovered / Cy* flies expected where Cy`flies expected = Cy flies / 2. Where two mutant allelescomplemented, Cy+ flies of both sexes were examined for fertilityand visible phenotypes.Fertility tests Fertility tests on viable mfs(2)31 hypomorphs were performedas follows. Parental crosses (mfs/CyO x mfs/CyO) were at either25° or 29°. The Cy' flies were collected within 12 hours ofeclosion and tested for fertility at either 22°, 25°, or 29° bymating with +/+ (Oregon-R) control males. Typically the flies weretested in groups of 25 with a minimum of 100 flies tested per sexand strain. Cultures were cleared of adults after seven days andwere then observed until all viable progeny had eclosed.Lethal phases The lethal phase was determined for each of the four mfs(2)31alleles. mfs(2)31 / + males were crossed to Df(2L)J77 / + femalesat 25°. All progeny from this cross should survive except thosewith the genotype mfs(2)31 / Df(2L)J77. Following a twelve houregg lay, eggs were collected, counted and transferred to freshvials. A minimum of five vials were established for each strainwith approximately 100 eggs per vial. Embryonic death was recordedat 72 hours; white eggs were recorded as unfertilized whiledarkened eggs were recorded as deaths. Larval death was determined72by failure to pupate while pupal death was determined by failure toeclose. Total deaths for a particular developmental stage werecompared to the percentage of deaths recorded in the appropriatecontrol crosses.Pigment Assays All flies used for pigment assays were raised at 22°, in 1/2pint milk bottles with a maximum 50 flies per bottle. Flies of thecorrect genotype were collected within 24 hours of eclosion, keptat 22° for 48 hours, and then frozen at -70° for at least 30minutes. A minimum of 25 flies per sex per genotype werecollected. Immediately after removing the flies from the freezer,they were decapitated by banging the vials on a pounding pad. Foreach genotype, five heads were put into the well of a microtiterplate (Nunc microtiter plates), 5 times for a total of 25 flies pergenotype. The heads were sonicated vigorously for five seconds in30 ul of sonicating solution (1% NH 4OH, 0.25 M (3-mercaptoethanol)with care taken to ensure that all heads remained immersed.Slides were prepared by attaching Whatman 3mm filter paper toone side with glue sticks providing adequate adhesion. Fivemicroliters of solution from each well was spotted onto a preparedslide (Whatman side) using a Gilson P20 pipetman. Slides wereprotected from bright lights and were not read until dry. Theamount of pigment in each spot was determined fluorometricallyusing an MPS-1 Zeiss microscope. The values obtained were averagedand expressed as a percentage of Oregon-R control flies.73Ovary dissections Virgin mfs 1 /mfs 1 females were collected from cultures which hadbeen developed at 25°.^The flies were held at 25° for theappropriate period of time prior to dissection.^Ovaries weredissected and photographed in Drosophila ringers (6.5 g NaC1, 0.14g KC1, 0.12 g CaC1 2 , 0.20 g NaHCO 3 , 0.01 g NaH 2 PO 4 in 1000 ml dH 2O;pH 7.2). Photographs were interpreted according to the drawings ofKing (1970).Thick sections of ovaries Freshly dissected ovaries from seven day old virgin females(25°) were fixed in glutaraldehyde (2.5% in 0.1 M cacodylatebuffer) for eight hours and then washed in cacodylate bufferovernight. Postfix was in 1% osmium tetroxide for 1 hour followedby a rinse in distilled water. Next, the tissue was seriallydehydrated on the following schedule: 10 min in 50% ethanol: 10 minin 70% ethanol: 10 min in 85% ethanol: 10 min in 95% ethanol: 2 X10 min in 100% ethanol. The ovaries were transferred to 50%propylene oxide in absolute ethanol for 30 minutes followed by 2 X30 minutes in 100% propylene oxide. Embedding was done serially onthe following schedule: 30 min in 50% spurr resin / 50% propyleneoxide: 30 min in 75% spurr resin / 25% propylene oxide: 30 min in100% spurr resin. Polymerization was in an oven at 60° for 16hours. Ovaries were sectioned and stained with toluidine blue (1%toluidine blue : 1% aqueous borax).74Reversion screen The mfs 2 allele was recovered in a P element screen (seeChapter 2). To determine whether or not the mutation was theresult of an insertional event, the allele was outcrossed underdysgenic conditions and then selected for viability in trans witha deficiency which uncovers the locus. The protocol for thisscreen for revertants can be found in Figure 6 . The mfs 2 b pr cnmales were crossed en masse to females carrying the 02-3 Sbchromosome, a potent source of transposase activity. Dysgenicmales carrying both the mfs 2 b pr cn chromosome and the A2-3 Sbchromosome were collected and crossed to Df(2L)J77 females. Agerm-line reversion of the mutation is indicated by the presence ofexceptional Cy' progeny. Such progeny have the genotype mfs 2 b prcn/Df(2L)J77, a normally lethal combination of second chromosomes.In situ hybridization to polytene chromosomes Black cells (Bc) is a dominant larval marker on the secondchromosome. Males carrying a Bc chromosome and an mfs2 b pr cnchromosome were crossed to wild-type virgin females at 17°. Latethird instar larvae carrying the mfs2 b pr cn chromosome (Bc') wereselected for in situ hybridizations. Salivary glands weredissected in Drosophila ringer, fixed for 45 seconds in 45% aceticacid and then squashed in 10 ul of a solution of 1 part lacticacid: 2 parts distilled water: 3 parts glacial acetic acid. Thechromosomes were allowed to flatten overnight at 4° and were thenflash frozen in liquid nitrogen. The coverslips were removed using75Figure 6: Screen for revertants of mfs(2)312.mfs2 b pr cn^+^Sp . 4a■ 2-3 SbXCyO + CyO^UbxJmfs2 b pr cn 46,2-3 Sb^Df(2L)J77 . +,+ CyO^+CyOmfs2 b pr cn^+Df(2L)J77^+REVERTANT76a sharp razorblade and the slides were immediately immersed in pre-chilled (-20°) 95% EtOH which was then allowed to come to roomtemperature. The slides were removed from the ethanol, air-driedand stored at 4° until needed.Immediately prior to incubation with a biotinylated probe, theslides were immersed for 10 minutes in 250 ml of 0.1 Mtriethanolamine-HC1 solution (pH 8.0) to which 310 ul of aceticanhydride had been added. The slides were then washed (2 X 5minutes) in 2 X SSC and dehydrated sequentially in 70% EtOH (2 X 5minutes) and 95% EtOH (2 X 5 minutes). The chromosomes were air-dried and then denatured in freshly prepared 0.07 M NaOH. Theywere again washed and sequentially dehydrated as described above.The slides were air-dried and the probe was added immediately.Nick-translated probes (1 ug of DNA) were prepared usingbiotin-11 dUTP (BRL) and the BRL Bionick labelling system accordingto the manufacturers instructions. Probes were purified by asingle ethanol precipitation and re-suspended in 117 ul water.They were denatured by boiling for 5 minutes and were then plungedinto ice water. The following was added to the denatured probe: 40ul 20 X SSC, 40 ul 10% dextran sulfate and 4 ul 50 X Denhardt's (50X Denhardt's reagent contains 5 g of Ficoll, 5 g ofpolyvinylpyrrolidone (PVP), 5 g of bovine serum albumin (BSA), andH 2O to 500 ml). Approximately 20 ul of probe was added to eachslide which was then covered with a small coverslip and sealed withrubber cement. Chromosomes were incubated for 12 hours in amoisture chamber at 58°. The coverslips were then removed and the77slides were washed sequentially as follows: 3 X 20 min in 2 X SSCat 53°, 2 X 10 min in 2 X SSC at room temperature.Hybridization was detected using the BRL Blugene kit and phasecontrast microscopy. Slides were incubated for 10 min in Buffer 1(0.1 M Tris (pH 7.5), 0.1 M NaCl, 2mM MgC1 2 , 0.05% Triton X-100(Sigma)) at room temperature and 20 min in Buffer 2 (Buffer 1 plus2% BSA) at 42°. The solution was moved to room temperature andallowed to sit for an additional 10 minutes. One hundred ul of asolution of 2 ul strepavidin-alkaline phosphatase conjugate in 1 mlBuffer 1 was added to each slide which were then incubated at roomtemperature for 2 hours. Slides were washed for 2 X 3 min inBuffer 1 and 1 X 3 min in Buffer 3 (0.1 M Tris (pH 9.5), 0.1 MNaC1, 50 mM MgC1 2 ). One hundred ul of a solution of 4.4 ul NBT and3.3 ul BCIP in 1 ml Buffer 3 was added to each slide. Slides werecovered with a large coverslip and incubated overnight in adarkened moisture chamber at room temperature. The coverslips wereremoved, the slides were washed briefly in distilled water and thechromosomes were examined under water with phase contrast optics.78ResultsComplementation of mfs(2)31 with Su(var) loci A number of dominant suppressors of PEV have been localized todivision 31, either by recombination or by deficiency mapping.Although mfs l exhibits no dominant suppression, it may be a weakallele of one of these previously described modifiers.Complementation tests were conducted between mfsl andrepresentative alleles of closely linked Su(var) loci, the resultsof which are summarized in Table 5. Transheterozygotes wereexamined for fertility and visible phenotypes.Su(var)212 is a strong dominant suppressor which maps to theright arm of chromosome 2 (Sinclair et al., 1983). Although itsmap position would indicate that it is not allelic to mfs(2)31, itwas included in the complementation tests as a control for generalinteractive phenotypes between mfs(2)31 and Su(var) loci.Su(var)212 and mfs l exhibited no phenotypes when placed in trans.Suvar(2)1 alleles show sex specific lethality and sterility intrans with Df(2L)J2 but fully complement Df(2L)J106, Df(2L)J27 andDf(2L)J77 (Sinclair et al., 1991). This pattern of complementationis consistent with the recent isolation of Suvar(2)1 alleles whichare associated with P element insertions in 31B-C (G. Reuter,personal communication). Heteroallelic combinations of Suvar(2)1exhibit a variety of phenotypes including eye discoloration anddefective wings. To test for allelism between mfs(2)31 andSuvar(2)1, mfs l was crossed to two of the stronger Suvar(2)1alleles: Su(var)213 and Su(var)214. Transheterozygotes were79TABLE 5Complementation with dominant Su(var)sCrossmfs 1/Cy0 x b it r1/Cy0mfs l /C_y0 x Su(var)212/Cy0Infs l /Cy0 x Su(var)213/Cy0mfs 1/Cy0 x Su(var)214/Cy0mfs l /Cy0 x Su(var)204/CyOmfs l/Cy0 x Su(var)205/Cy0mfs 1/Cy0 x Sup20/Cy0mfs l /Cy0 x Sup34/CyOmfs 1/C_y0 x Sup37/CyOmfs l /Cy0 x Sup39/CyOmfs l/C_y0 x Sup41/Cy0mfs l /Cy0 x Sup46/Cy0mfs l /Cy0 x Sup47/CyOViabilitya FertilityMale Female0.98 (381) Fertile Fertile1.17 (766) Fertile Fertile0.80 (642) Fertile Fertile0.92 (1132) Fertile Fertile0.76 (999) Fertile Fertile0.97 (880) Fertile Fertile0.81 (246) NDb Fertile0.87 (205) ND Fertile0.76 (185) ND Fertile0.86 (297) ND Fertile0.86 (408) ND Fertile0.85 (292) ND Fertile1.08 (277) ND Fertilea Viability = (Cy* progeny recovered) / (Cy* progeny expected)from the cross mfs l /Cy0 X Su(var)/Cy0. Total progeny scored areprovided in parentheses.b ND = Not done80counted, examined for visible phenotypes and tested for fertility.Based on all these criteria, mfs(2)31 fully complemented Suvar(2)1.Su(var)204 has been mapped by recombination to division 31(Sinclair et al., 1991). Since this mutation is homozygous viableand fertile and shows no phenotype over deficiencies for theregion, its precise cytological location is not known. As is thecase with Suvar(2)1, transheterozygotes of mfs(2)31 and Su(var)204were completely viable and fertile, showing no visible phenotypes.Su(var)205 has been cloned and localized to cytologicalsubdivision 29A. This gene encodes the non-histone chromosomalprotein HP1, a protein which specifically associates with division31 on salivary gland polytene chromosomes (James et al., 1989).Although Su(var)205 is not allelic to mfs(2)31, the possibilityexists that it regulates the expression of genes within division 31and thus may exhibit a phenotype in trans with mfs(2)31 alleles.A complementation test between these two mutations did not provideany evidence of such an interaction.Sup20, Sup34, Sup37, Sup39, Sup41, Sup46, and Sup47 are weakdominant suppressors of PEV which map by recombination to division31 (G. Reuter, personal communication). These are homozygousviable strains which have not been tested for allelism nor havethey been tested against division 31 deficiencies. Strains whichwere heterozygous for these mutations and the mfs' chromosome wereviable and exhibited no visible phenotypes. The femaletransheterozygotes were all fertile but the males were not tested.To summarize, mfs(2)31 was tested for allelism with all81available dominant suppressors of PEV which have been localized todivision 31. Using viability, male and/or female fertility, andvisible phenotypes as a criteria, mfs(2)31 does not appear to beallelic to any previously described Su(var) loci.mfs(2)31 inter se complementation Hypomorphic alleles can be useful for determining a genesfunction, particularly if they exhibit phenotypes at well definedstages in development or in easily recognizable tissues.Similarly, alleles which represent complete loss-of function at alocus can sometimes be used to determine the earliest stage atwhich a gene product acts. In order to determine the relativestrengths of the four mfs(2)31 alleles and identify viable alleliccombinations, a complementation matrix was elaborated for themfs(2)31 locus (Table 6). In no case did the parental source ofthe mutation influence viability. Similarly, no differences inviability were observed if the crosses were done both at 25° and at29°.All mfs(2)31 alleles were recovered on the basis of theirfailure to complement a deficiency for division 31 and all showedcomplete lethality when tested on a larger scale over Df(2L)J77(see Table 6). Although rare survivors of mfs i in trans withDf(2L)J106 have been previously described (Sinclair et al., 1991;D. Sinclair, personal communication), no survivors over Df(2L)J77were recovered in 10,000 flies scored.The only mfs(2)31 allele which survived as a homozygote was82TABLE 6Complementation matrix for mfs(2)31'mfs 1 mfs2 mfs3 mfs4DfJ77 29° 0.00 (4721) 0.00 (2763) 0.00 (3714) 0.00 (1527)25° 0.00 (5716) 0.00 (1825) 0.00 (1165) 0.00 (1450)mfs 1 29° 0.50 (2658) 0.73 (4896) 0.00 (1651) 0.00 (2110)25° 0.48 (1684) 0.68 (3331) 0.00 (1944) 0.00 (1794)mfs2 29° 0.00 (1700) 0.00 (2768) 0.00 (1231)25° 0.00 (1296) 0.00 (2368) 0.00 (1138)mfs 3 29° 0.00 (878) 0.00 (1106)25° 0.00 (910) 0.00 (1230)mfs 4 29° 0.00 (182)25° 0.00 (423)a Data is recorded as (Cy' progeny recovered)/(C/progeny expected)from the cross mfs/CyO X mfs/CyO. Total number of flies scoredis provided in parentheses.83the original mfs 1 mutation. Homozygotes of both sexes exhibitedthe small, slender bristles and sterility previously described bySandler (1977). These flies were recovered at a frequency of 50%of expected, which is substantially higher than the previouslyreported value of 25% (Lindsley et al., 1980). This discrepancymay be attributable to culture conditions, however it is possiblethat a portion of the homozygous progeny were missed in previousdeterminations of viability, since mfs 1 homozygotes have a veryprolonged developmental period, eclosing several days later thantheir heterozygote siblings. The other alleles of mfs(2)31 werecompletely lethal both as homozygotes and as hemizygotes. The mfs 4strain has a mild reduction in fertility in both sexes although itcan be maintained as a balanced stock. The reduced fertility isreflected in the relatively low number of heterozygotes recoveredin the test for homozygous viability.The only other viable combination of mfs(2)31 alleles wasrlfs l /mfs 2. Like the mfs 1 homozygotes, these flies have shortenedthoracic bristles in both sexes and are male sterile. They differfrom mfs 1 /mfs 1 flies in that the females are fertile and thefrequency of recovery is somewhat higher: 0.68 vs 0.48 at 25°.A reasonable interpretation of this complementation matrix isthat mfs l and mfs 2 are hypomorphic alleles while mfs 3 and mfs 4approximate the complete loss-of-function phenotype at the mfs(2)31locus. Since all the alleles are lethal as hemizygotes, and threeare lethal as homozygotes, the gene would appear to have anessential function. Although mfs 1 is the only allele to survive as84a homozygote, it is difficult to judge the relative strengths ofmfs 1 and mfs 2. Compared to the mfs l homozygotes, the mfs l /mfs 2allelic combination survives better and has less severe fertilityproblems. It is plausible that mfs 2 is the weaker allele and onlyfails to survive as a homozygote because of secondary lesions.Fertility of mfs(2)31 alleles The failure of mfs(2)31 hypomorphs to elaborate wild-typebristles coupled with their severe fertility problems suggests thatthe gene product may function late in development or in adulthood.It has been reported that the sterility associated with mfs 1homozygotes is temperature-sensitive (Lindsley et al., 1980)although the nature of the sensitivity has not been well defined.If there is an adult temperature-sensitive period for sterility, itwould indicate a germ-line associated adult function for the geneproduct.Temperature shift experiments were conducted on mfs(2)31hypomorphs and fertility was examined under a variety oftemperature regimes. Flies were reared at either 25° or 29°,collected within 24 hours of eclosion and tested for fertility ateither 22°, 25° or 29°. This protocol made it possible todistinguish between sensitivity due to developmental temperatureand sensitivity due to adult temperature. The results of thisstudy are summarized in Table 7.Regardless of the temperature regime, all males that weretested were sterile, thus providing no evidence for, or against, an85TABLE 7Fertility of mfs(2)31 hypomorphsGenotype' Temp.RearedTemp.TestedMales Femalesmf,s 1/mfs 1 29° 29° Sterile Sterile;Lay no eggs25° Sterile Sterile;Lay no eggs22° Sterile Sterile;1 or 2 eggs25° 29° Sterile Sterile;1 or 2 eggs25° Sterile Sterile;1 or 2 eggs22° Sterile sterile;small,^egg laymfs 1/mfs 2 29° 29° Sterile Sterile;large,^egg lay25° Sterile Semi-sterile;1 or 2 eggs hatch anddevelop into adults22° Sterile Semi-sterile;1 or 2 eggs hatch anddevelop into adults25° 29° Sterile Sterile;Large,^egg lay;Eggs hatch but die as1st instar larvae;A few pupate but noeclosion25° Sterile Fertile22° Sterile Fertilea The mfs 1 allele was induced on an Oregon-R chromosome while mfs 2is on a b pr cn chromosome.86adult function for the gene product. The results of the femalefertility tests were somewhat more informative. In the case ofmfs 1 homozygotes, females were sterile under all conditions testedalthough they differed in their ability to deposit eggs. Theability to lay eggs depended both on the temperature at which theflies were reared and the temperature at which they were tested.Thus, if flies were reared at 29° and tested at 22° they onlyproduced one or two eggs per vial (from 25 females), whereas ifthey were reared at 25° and tested at 22° they produced 50-100 eggsper vial. A similar effect was observed if the adult temperaturewas varied. Thus, flies reared at 25° and tested at 29° producedonly one or two eggs whereas flies reared at 25° and tested at 22°produced 50-100 eggs per vial. The control strain (Oregon-R) wascompletely fertile under all temperature regimes. It would appearthat the ability of the sterile females to deposit eggs issensitive both to developmental temperature and the temperature ofadult cultures. Interestingly, egg deposition in mfs 1 homozygotesand mfs l /mfs 2 heteroallelic females resembles the lethalityphenotype.The pattern of temperature-sensitivity was also observed inthe fertility tests of mfs 1 /mfs 2 heteroallelic females. Theinfs 1 /mfs 2 females were highly fecund, that is they producedsubstantial amounts of eggs regardless of the temperature regimeexamined. However, the ability of the eggs to hatch and developinto adults depended both on developmental temperature and thetemperature at which the adults were maintained. Thus, flies87reared at 29° and tested at 22° produced only one or two eggs whichwere able to hatch and develop into adults while flies reared at25° and tested at 22° were completely fertile. The fertilityphenotype also varied as a consequence of temperature at whichadults were maintained. Flies which were reared at 25° and testedat 29° were completely infertile while siblings which were testedat 22° were completely fertile. The control strain (b pr cn/+) wasfertile under all temperature regimes.To summarize, the degree of sterility of mfs(2)31 hypomorphs(females only) is sensitive both to the temperature at which theflies are developed and the temperature at which they aremaintained as adults. This latter observation indicates that thegerm-line associated function of the mfs(2)31 gene productcontinues into adulthood. The more severe sterility observed inthe mfs 1 homozygotes is correlated with a lack of oviposition infemales.Studies on mfs(2)31 oogenesis The sterility phenotype of the mfs 1 homozygotes is associatedwith a failure to deposit eggs. Whether the lack of fecundityreflects a failure in one of the systems controlling oviposition oran inability of the flies to produce mature eggs is not clear.However, the lengthy temperature-sensitive period favors the latterhypothesis. Indeed, all elements of the reproductive system, withthe exception of the ovaries, are morphologically normal inmfs l/mfs 1 females raised at 29°. Thus, a defect in one of these88structures is an unlikely source of the sterility phenotype. Incontrast, oogenesis is an ongoing process in the adult involving abalance and interaction between a number of regulatory factors (forreview see Mahowald and Kambysellis, 1980). Since the fidelity ofthese adult processes is of utmost importance in controlling thetiming and rate of egg maturation and deposition, defects in anumber of these processes may exhibit a sensitivity to temperature.As a preliminary step in the investigation of the reproductivedefect in mfs 1 homozygote females, the temporal pattern of ovariandevelopment was examined in adults. Ovaries were dissected frommfs l /mfs l virgins at 3 hours, 12 hours, 24 hours, 48 hours, 72hours, 96 hours, 10 days and 21 days post-eclosion and compared toovaries from similarly staged Oregon-R females. Representativeovaries from the various stages are shown in Figure 7.At 3 hours post-eclosion (plate A), ovaries from mfs(2)31 andOregon-R females are virtually indistinguishable. The vitellariumsappear normal in number and morphology as do the lateral and commonoviducts (see arrow). As is expected for this stage indevelopment, there is no evidence of vitellogenesis in either themutant or wild-type ovaries.Over the next 24 hours (plates b and c), the morphology ofmfs 1 /mfs 1 and Oregon-R ovaries becomes strikingly different. In thewild-type ovaries, vitellogenesis commences and many stage 9-11oocytes are identifiable. In contrast, there is no indication thatvitellogenesis is occurring at this time in mfs 1 homozygotes. Theovaries of mfs 1 homozygotes become moderately enlarged, but this89Figure 7: Ovarian development in mfs l/mfs l adults raised at 25°.Ovary(s) on left of plates are from mfs l /mf,s 1 whileovary(s) on right are from Oregon -R. Plates E and Fshow only a single wild-type ovary for comparison whileplates G and H contain only single ovaries from bothstrains. Plate A: 3 hrs post-eclosion (arrow indicatesintact common and lateral oviducts), Plate B: 12 hrspost-eclosion, Plate C: 24 hrs post-eclosion, Plate D:48 hrs post-eclosion, Plate E: 72 hrs post-eclosion,Plate F: 96 hrs post-eclosion (arrow indicates stage 10-12 ovarian cyst), Plate G: 10 days post-eclosion, PlateH: 3 weeks post eclosion. Plate I: Thick section through96 hr mfs l /mfs l ovary showing a stage 9 ovarian cyst.9091appears to be due to the accumulation of oocytes at the pre-vitellogenic stages.By 72 hours of development (plates D and E) the difference inmorphology between wild-type and mutant ovaries has become evenmore pronounced. The ovaries of wild-type flies are completelymature; they contain many fully-developed eggs. In contrast, themfs 1 ovaries appear arrested in development; the eggs appear to beessentially pre-vitellogenic. However, some yolk deposition maybeoccurring in the mutant ovaries by 72 hours; there are a few opaquestage 8 oocytes (plate E).At 96 hours (plate F), vitellogenesis in mfs l ovaries is moreapparent. At this time, it is not uncommon to see individualovarioles within which vitellogenesis has proceeded as far asstages 10-12 (see arrow in plate F) in the mutants. However, forthe most part, vitellogenesis is proceeding very slowly and isstill not evident in the majority of the ovarioles.By day ten (plate G), the process of vitellogenesis hasadvanced in the mutant ovaries and eggs in the latter stages ofdevelopment are present in virtually all the ovarioles. Althoughthe wild-type virgins have commenced egg deposition, the mfs 1 Amfs 1females do not lay eggs at this time.By three weeks (plate H), the mfs l homozygotes have still onlydeposited 1 or 2 eggs and the morphology of the ovaries has changedonce again with respect to wild-type ovaries. Although fullyformed eggs are clearly visible, there is no evidence of pre-stage12 oocytes. This is presumably due to the reabsorption of these92oocytes, a phenomenon which has been described in Drosophilafemales which fail to undergo oviposition (Wilson, 1985; Giorgi andDeri, 1976).The yolk proteins in developing Drosophila oocytes originatefrom two sources, the follicle cells and the cells of the fatbodies (Mahowald and Kambysellis, 1980; Brennan et al., 1982).Vitellogins from the fat bodies are released into the hemolymphwhere they are transported to the oocytes and taken up bypinocytosis. Thus, failure to undergo vitellogenesis may reflecta defect in the fat bodies or follicle cells. Ovaries from 96 hourmfs 1 homozygotes were sectioned and examined. A representativesection through a stage 9 oocyte (the first stage in which yolkdeposition is detectable) is shown in plate I (Figure 7).In all ovarian sections examined, no defects were detected atany stage in oogenesis. In particular, the follicle cell, nursecell and oocyte morphology appear to be normal at the onset ofvitellogenesis (stage 9). By stage 9 of oogenesis, the majority ofthe follicle cells have moved over the rapidly growing oocyte andassumed a columnar morphology, while the remainder have become thinand squamous covering the 15 nurse cells. As can be seen in Figure7, the follicle cells of mfsi homozygotes have proliferated,migrated properly and assumed a wild-type morphology. The nucleiare visible and positioned normally in the columnar follicle cells.The border cells, a group of anterior follicle cells which migratesas a cluster to the anterior surface of the oocyte, aremorphologically normal and correctly positioned in the mfs 193homozygotes.The correct number of nurse cells are present in the mfs 1ovarian cysts and their orientation and size with respect to theoocyte appears normal. As is the case with stage 9 wild-type nursecells, polyploid chromosomes are easily distinguished, dispersedthroughout the nurse cell nuclei.Within the oocyte cytoplasm, protein yolk spheres are observedat a density equivalent to similarly staged wild-type oocytes. Asexpected, the oocyte nucleus has migrated to a position adjacent tothe nurse cells at the future antero-dorsal position in the egg.To summarize, the sterility in mfs 1 homozygotes is associatedwith a delay in the onset of vitellogenesis. Yolk deposition isfirst detected 72 hours following eclosion and the ovaries do notachieve full maturity for an additional six days. Although fullymature eggs are detected in the ovaries after 10 days, and themorphology of the common and lateral oviducts appears normal,oviposition is rare. It is not clear whether the delay invitellogenesis and the failure to oviposit are the consequence ofthe same or independent events. Thick sections through ovaries ofmfs 1 homozygotes revealed no obvious morphological defects at anystage in ovarian development.Lethal phases for mfs(2)31 In addition to an adult germ-line associated function, themfs(2)31 product acts early in development as evidenced by thefailure of strong alleles to survive to adulthood. Strong alleles94of essential loci can be useful for determining the earliest stageat which the essential function is occurring. Weaker alleles canalso be useful for defining the function of a gene product,particularly if they die at well-defined stages in development i.e.the time of elaboration of a particular tissue. As part on anongoing analysis of the various functions of the mfs(2)31 locus,the lethal phase was determined for all four mutant alleles. Thealleles were tested in trans with Df(2L)J77 and the crosses weredesigned such that only the mfs(2)31/Df(2L)J77 progeny should die.The results of the lethal phase analysis are summarized inTable 8. The Oregon-R chromosome is the control for mfs 1 while bpr cn is the control for the remainder of the alleles. These arethe chromosomes upon which these mutations were induced. Thecrosses were done in one direction only using Df(2L)J77 females,and thus the results should not reflect any differences in allelespecific maternal effects. All four alleles exhibit one principlephase of lethality during larval development. Although mfs 1 andmfs 2 appear to be weaker alleles based on their pattern ofcomplementation, this is not reflected in the lethal phaseanalysis.mfs(2)31 alleles suppress position -effect variegation It has been previously reported that mfs' homozygotes suppressthe white (w) variegation associated with the w' rearrangement, aphenotype which has also been observed in rare mfs 1 /Df(2L)J106survivors of both sexes (Sinclair et al., 1991). Although this95TABLE 8Lethal phases for mfs(2)31 alleles 4Strain TotalScored%EmbryonicLethalityLarval PupalOregon-R 505 0.0 4.0 2.7b pr cn 594 2.7 8.0 7.8mfs l 585 2.7 35.5 4.6mfs 2 b pr cn 457 3.9 39.6 4.8mfs 3 b pr cn 691 2.3 25.3 8.5mfs 4 b pr on 849 4.5 25.0 7.0a All chromosomes were tested over Df(2L)J77.96mutation exhibits no dominant phenotype with respect to PEV,interallelic complementation indicates that it is a weak allele.This raises the possibility that stronger alleles such as mfs 3 andmfs 4 may exhibit dominant phenotypes. In addition, since only onemfs(2)31 allele has been tested for suppression, it is formallypossible that the su(var) phenotype is a consequence of a tightlylinked second site mutation which is also uncovered by Df(2L)J106.Alternatively, the phenotype could be a peculiarity associated withthe one mfs(2)31 allele, if for example it is acting as anantimorph or neomorph with respect to suppression. Gain-of-function alleles have been previously described for other Su(var)loci (Sinclair et al., 1983; 1991). In order to address all ofthese issues, the four mfs(2)31 alleles were tested for dominantand recessive (where testable) suppression of PEV in a variety ofvariegating backgrounds. The results of this study are summarizedin Table 9.The four mfs(2)31 alleles were tested for their ability tosuppress the white variegation associated with wino wrn 5 1 and W'.These are X-chromosome inversions with proximal breakpoints in thecentric heterochromatin. Since the distal breakpoints in wm 5 1 b andW' are distal to the white gene, these rearrangements move thevariegating locus to the variegation inducing region. In contrast,the distal breakpoint of DV is proximal to the white gene thusmoving the variegation inducing region to the affected locus.The W' and vin' b strains rapidly accumulate modifiers inbalanced stocks becoming such weak variegators that they are of97TABLE 9Pigment assaysSex Pigment Levels'*wm 4^5.7^±^2^25.3^±^16.3^±^221.2^± 2^7.1^±^1wm 5 1 b31. ^± 418.9^±^234.2^± 414.2^±^139.3^±^3Wnj14.764.718.567.022.4± 4± 6± 3± 5± 2V 25.4 ± 3 22.8 ± 2 66.7 ± 56.1 ± 1 34.7 ± 4 21.5 ± 2g 30.5 ± 7 11.3 ± 1 69.1 ± 48.3 ± 1 37.7 ± 2 15.9 ± 113.1 ± 5 22.9 ± 4 65.4 ± 6e 11.1 ± 2 31.2 ± 5 17.3 ± 1V 29.7 ± 5 20.9 ± 2 74.4 ± 6a' 66.5 ± 9 nrb 97.1 ± 289.7 ± 5 45.9 ± 9 90.8 ± 6e 50.0 ± 3 64.8 ± 8 44.5 ± 7V 28.7 ± 4 47.6 ± 3 68.2 ± 4Strainb pr cn / +mfs 1 / +mfs2 b pr cn / +mfs 3 b pr cn / +mfs 4 b pr cn / +mfs' / mfs lmfs' / mfs 2 b pr cna Pigment levels are expressed as a percentage of Oregon-R. Themeasurements of tslb and W' females are for Te5m/w- andW'/w- flies respectively. All measurements were done at 22°.b No flies recovered.98little use for assaying suppression. Because of this, chromosomeswere used which had been outcrossed for seventeen generations to wmales, and which had been selected at every generation for strongvariegation. To further reduce the pigment levels and thus make amore sensitive assay, pigment measurements were made on wm 51b/w- andW'/w- females rather than w'/w' and ves /winsibControl pigment levels for the mfs 1/+ and mfs 1 tmfs 1 strains arefrom a wild-type strain while levels for mfs2/+, mfs 3 /+, mfs 4/+ andmfs 1 /mfs 2 are from a b pr cn/+ strain. When compared to theirrespective controls, none of the mfs(2)31 alleles exhibit dominantsuppression in any of the rearrangements examined. In contrast,recessive suppression is associated with the mfs 1 /mfs 1 and mfs 2 /mfs 2strains in all three variegating backgrounds. The phenotype isparticularly strong in the mfs 1 homozygotes where near wild-typepigment levels are observed in le and W flies of both sexes anda moderate increase in pigment (twofold) is observed in wm 51bfemales. No winsib/y ; mfs 1 /mfs 1 males were recovered in this study.Although substantial and consistent, the suppressionassociated with the mrs 2 /mrs 2 strain is more moderate than in themfs 1 homozygotes. Increased pigment levels on the order ofmagnitude of 25%-45% of Oregon-R (2-10 fold increases in pigmentlevel) are observed for males in all three variegating backgroundsas well as in W" 51b females. However, no suppression is observed inW' and w" females. This is not surprising since this alleliccombination has a weak phenotype for female fertility.The isolation of a second mfs(2)31 hypomorphic allele with a99su(var) phenotype confirms the initial observation of Sinclair etal. (1991) that lesions at the mfs(2)31 locus are associated withsuppression of PEV. Since these mutations have been isolated onthe basis of lethality rather than suppression, it is unlikely thatthey are both gain-of-function alleles. It is more reasonable toconclude that loss-of-function at the mfs(2)31 locus produces asu(var) phenotype. The absence of a dominant phenotype in even thestrongest alleles would indicate that the gene product is not asdosage-sensitive as are the majority of previously describeddominant modifiers of PEV. Finally, the su(var) phenotype does notappear to be due to a specific interaction between mfs(2)31 and thew" inversion since suppression has been observed in a variety ofvariegating backgrounds.mfs 2 has a P element in 31E The mfs2 mutation is a weak allele of mfs(2)31 which wasisolated in a P element screen. A P element probe was hybridizedto salivary gland polytene chromosomes from the mfs 2 strain. Thein situ hybridization revealed an element in subdivision 31E onchromosome arm 2L (Figure 8), the known cytogenetic location ofmfs(2)31 (see Chapter 2).Reversion of mfs 2 P element insertions often produce weak mutant phenotypes.However, it has been previously reported that a substantialproportion of mutations induced in P element screens are100Figure 8: In situ hybridization to salivary gland polytenechromosomes of a mfs2 b pr cn/+ strain. A P elementspecific probe reveals homology in cytologicalsubdivision 31E (arrow).101attributable to chromosomal rearrangements. It is possible todistinguish between these two classes of mutation by out-crossingthe mutant chromosome in a dysgenic background and measuring thefrequency of reversion of the mutant phenotype. Under dysgenicconditions, reversion frequencies are considerably higher forinsertional mutations than for mutations caused by rearrangements.The 02-3 Sb chromosome is a potent source of transposaseactivity which was used to revert the lethal phenotype associatedwith mfs2/Df(2L)J77 transheterozygotes (see Figure 6). A total of2500 chromosomes were examined from which nine exceptionalmfs2/Df(2L)J77 flies were identified. This represents a reversionfrequency of over 1% which is at least 5 orders of magnitude higherthan spontaneous reversion frequencies (Ashburner, 1989). Allrevertants were recovered from different bottles and are likely tohave arisen from independent events. The putative revertantchromosomes were established in balanced stocks and all ninecomplemented Df(2L)J77 and mfs 4 on subsequent retests. Theserevertants are described in Table 10.Of the nine revertants recovered, five were male and four werefemale. Three of the females and one of the males were Sbindicating that they still carried the A2-3 Sb chromosome. Therevertants were tested as homozygotes and all were viable andfertile in both sexes. This strongly argues against the mfs 2chromosome carrying any second site lethal and/or sterilemutations. Stocks established from female and/or Sb revertantshave not been used for any subsequent analyses because of the102TABLE 10Revertant descriptionRevertant Strain^Sex^Stubble^Homozygotesmf I'^s b pr cn^ e^+^viable/fertilemf '^s b pr cn e^+^viable/fertile^mfs" b pr cn^ e^+^viable/fertile^mfsR4s b pr cn e^+^viable/fertile^mfsRS b pr cn^ e^-^viable/fertilemf R6^s b pr cn 9^+^viable/fertile^mfs R7 b pr cn^ 9^-^viable/fertile^mfs" b pr cn 9^-^viable/fertile^mfs" b pr cn^ 9^-^viable/fertile103possibility of unwanted recombinational and transpositional events.Three of the stocks established from male Sb' revertants wereselected for further analysis: lilts', mfsR2 and mfe3 . All threerevertant chromosomes were placed in a DP background and examinedfor bristle, sterility and su(var) phenotypes. The revertantchromosomes were tested as homozygotes, as heterozygotes with awild-type chromosome and in trans with the mfs 1 and mfs2 alleles.The results of these crosses are summarized in Table 11.When tested in trans with a wild-type chromosome the threerevertants have normal thoracic bristles and are fertile in bothsexes. Although all three have eye pigment levels at least as lowas control levels (b pr cn/+), the levels in mfsRl females areparticularly low showing an enhanced effect (4.2% vs 21.2% in thecontrols).None of the revertant chromosomes exhibit any aberrantphenotypes when examined as homozygotes. Although the pigmentlevels for these strains are very low, this is also the case forthe control strain (b pr cn/b pr cn) and probably reflects thepresence of two other mutations on the chromosome which influencepigment deposition, pr and cn.When crossed to mfs 1 and mfs2, the revertant chromosomes actas if the mfs 2 mutation is no longer present. The revertantssurvive over mfs2 and exhibit no abnormal phenotypes. Over mfs 1 ,they show none of the phenotypes associated with mfs 1 tmfs 2transheterozygotes: male suppression, slender bristles and malesterility. The complete reversion of the mfs(2)31 phenotype is104TABLE 11Revertant analysisStrain Sex Pigment(1,171114 ) aBristle Fertilityb pr cn/+ e 6.3^±^2 + Fertile9 21.2^± 2 + Fertileb pr cn/b pr cn e 6.3^±^2 + Fertile9 8.2^± 1 + FertilemfsRlb pr cn/+ e 2.4^±^1 + Fertile9 4.2^±^1 + FertilemfsR2b pr cn/+ e 3.1^±^1 + Fertile9 12.1^± 4 + FertilemfsR3b pr cn/+ e 6.8^±^1 + Fertile$^22.1^±^5 + Fertilemfelb pr cn/mfsRlb pr cn e 3.8^±^1 + Fertile9 2.6^±^1 + FertilemfsR2b pr cn/mfsR2b pr cn e 3.7^±^1 + Fertile9 3.9^±^1 + FertilemfsR3b pr cn/mfs R3b pr cn e 2.8^±^1 + Fertile9 3.7^±^1 + FertilemfsR lb pr cn/mfs 2b pr cn e 3.6^±^2 + Fertile9 2.6^± 1 + FertilemfsR2b pr cn/mfs 2b pr cn di 1.5^±^1 + Fertile9 5.1^±^1 + FertilemfsR3b pr cn/mfs 2b pr cn di 3.6^±^1 + Fertile9 4.0^± 2 + Fertilemfe lb pr cn/mfs 1 e 6.4^±^1 + Fertile9 15.5 ± 1 + FertilemfsR2b pr cn/mfs l e 8.4^± 1 + Fertile9 11.5^±^1 + FertilemfsR3b pr cn/mfs l e 6.9^±^1 + Fertile9 21.1^±^3 + Fertilea Pigment levels are expressed as a percentage of Oregon-R.105illustrated in Figure 9.In no case examined are the various mfs(2)31 mutant phenotypesgenetically separable; this provides compelling evidence that allthe phenotypes are a consequence of a single lesion at the mfs(2)31locus. Removing the P element insert from the mfs 2 allele revertsall the phenotypes associated with the mfs(2)31 locus. The highfrequency of reversion of mfs 2 in a dysgenic background indicatesthat this allele results from a P element insertion.106Figure 9: Reversion of su(var) and bristle phenotypes in a hybrid-dysgenesis induced revertant of mfs2 .Photographs are of adult males at 48 hrs post-eclosion:^A. wm4 /Y; +/+B. wm4/Y; mfs 1/mfs 1 C. wm4/Y; mfs 1/mfs 2 b pr cn D. w m4/Y; mfs 1/mfsR 1 b pr cn. .DiscussionThis chapter contains a phenotypic analysis of mfs(2)31, asuppressor of PEV which has been localized to cytologicalsubdivision 31E on the left arm of chromosome 2. Since division 31is known to contain one or more additional Su(var) loci that areboth haplo- and triplo-abnormal (Sinclair et al., 1991), it isdifficult to assess the nature of the mfs(2)31 mutation usingduplications and deficiencies alone. For example, although the fewduplications which are available for division 31 all show strongE(var) phenotypes, this cannot be directly attributed to themfs(2)31 locus since these duplications also encompass severalknown dosage-dependant modifiers of PEV (Sinclair et al., 1991).The deficiencies for 31E are somewhat more informative. Largedeficiencies which extend distally into 31A exhibit strongsuppression, presumably because they uncover Suvar(2)1, a haplo-dependant modifier which has been localized to 31B-C (Reuter etal., 1982; Sinclair et al., 1991). Smaller deficiencies whichuncover 31D-E show moderate suppression while deficiencies whichextend from 31D through 31F do not suppress at all (Sinclair etal., 1991). This pattern indicates that a moderately strong haplo-insufficient Su(var) locus is located in 31E, while actingepistatically to it, is an E(var) locus in 31F. In this regard, astrong E(var) strain which contains a P element inserted into 31Fwas recently isolated (G. Reuter, personal communication).This current study does not indicate that the moderate hapfo-insufficient Su(var) locus in 31D-E is mfs(2)31. Although complete108loss-of-function has not been formally demonstrated for any of themfs(2)31 alleles, none of them exhibit any dominant suppression ofPEV, regardless of the variegating background in which they areexamined. Typically, modification of PEV is a dominant phenotype,a feature which reflects a bias in the method of isolatingmutations rather than a functional characteristic of all modifyingloci. The mfs(2)31 locus differs from all previously describedmodifiers in that suppression is only observed in viablecombinations of hypomorphic alleles. There is some variability inthe phenotype which may be attributable to intrinsic properties ofthe variegators and/or the individual alleles.Although the data does not indicate that the mfs(2)31 gene ishaplo-insufficient with respect to modifying PEV, suppression doesappear to be associated with loss-of-function at this locus. Inorder to explain the sensitivity of the variegating phenotype tothe dosage of many Su(var) loci, it has been proposed that thesegenes encode components of multimeric complexes whose assembly isrequired for the formation of heterochromatin (Locke et al., 1988).The absence of dominant phenotypes for the mfs(2)31 allelessuggests that, unlike the previously described modifiers of PEV,the dosage of the gene product is not rate-limiting with respect tochromatin assembly. However, the mfs(2)31 gene product has anessential function, and weak alleles of mfs(2)31 have a strongsu(var) phenotype, suggesting that mfs(2)31 may have an importantrole in the chromatin assembly process.The effective lethal phase for mfs(2)31 alleles occurs during109larval development. The larval period is a time of rapid andsubstantial cell growth in Drosophila, processes which arefacillitated by polyploidization of most larval tissues. Althoughthere is no direct evidence that the mfs(2)31 gene product isrequired for protein synthesis and/or cell growth, there arephenotypic similarities between mfs(2)31 and mutant strains whichare known to be defective in these processes. Minutes are a classof haplo-insufficient mutation which are characterized by aprolonged developmental period and short, thin bristles (for reviewsee Kay and Jacobs-Lorena, 1987). These phenotypes are alsoobserved in viable mfs(2)31 allelic combinations. Molecularevidence suggests that the Minute loci are genes which encoderibosomal proteins and that the phenotypes reflect failures duringprocesses which require maximum rates of protein synthesis (Kay andJacobs-Lorena, 1987).Another process during Drosophila development which requireshigh rates of protein synthesis, and which is defective in mfs(2)31mutations, is the deposition of yolk proteins duringvitellogenesis. Females which carry viable mfs(2)31 alleliccombinations exhibit temperature-sensitive sterility with asensitive period extending into adulthood. The defect inoogenesis, as revealed by a temporal study of ovarian development,is associated with a delay in the onset of yolk deposition.Vitellogenesis does not begin in adult mfs 1 homozygotes until 72hours post-eclosion and mature eggs are not observed in the ovariesfor an additional seven days. In contrast, the wild-type ovary110reaches full maturity within 48 hrs of eclosion. The mutant straindoes not undergo any oviposition and the developing oocytes areeventually reabsorbed. The relationship between oviposition andvitellogenesis in Drosophila is unclear and it is difficult to saywhether the defects in these two processes is a consequence of thesame or independent events.Yolk protein synthesis in Drosophila occurs in both thefollicle cells and the fat bodies, subsequent to thepolyploidization of these respective cell populations (Mahowald andKambysellis, 1980). The follicle cells produce a substantialfraction of the yolk proteins, transporting it directly into therapidly growing oocyte (Brennan et al., 1981). I have examinedthick sections through stage 9 ovarian cysts and found that thefollicle cells of mfs 1 homozygotes have a wild-type morphology atthe onset of vitellogenesis indicating that there has not been ageneral failure on the part of these cells to proliferate andmigrate properly.Most of the yolk proteins are synthesized in the fat bodies,released into the hemolymph and transported into the oocytes by thepinocytotic activity of the oocyte itself (Mahowald andKambysellis, 1980). These processes are under hormonal controlwith the Drosophila juvenile hormone (JH) being the primarycontrolling agent (Postlethwait et al., 1976; Kambysellis and Heed,1974). Although the role that the mfs(2)31 gene product plays invitellogenesis is not known, mutations in genes such as apterous(ap), whose products are required for the final stages of111histolysis of the fat bodies as well as the uptake of yolk proteinsby the oocyte have similar phenotypes to mfs(2)31 (Postlethwait andWeiser, 1973). Flies which are homozygous for ap mutations aremale and female sterile and are defective in bristle elaboration.The female sterility is associated with a failure to deposit yolkin the developing oocyte.Although it is tempting to conclude from the bristle andovarian phenotypes that mfs(2)31 mutant strains are unable toachieve maximum rates of protein synthesis, it is difficult toreconcile such a developmental defect with the observedabnormalities in centriole migration (Lindsley et al., 1980) andthe suppression of PEV (Sinclair et al., 1991; this study). Thereis no evidence linking centriole behaviour with rates of proteinsynthesis while a low rate of protein synthesis would enhancerather than suppress PEV. The centriole phenotype could beexplained if centriole migration is a prerequisite forpolyploidization in Drosophila. A failure of either the folliclecells to achieve full polyploidy could slow down or delay yolkprotein synthesis and its subsequent deposition. Polyploidy isalso employed as a mechanism during larval development forsustaining rapid and substantial cell growth and as a mechanism forthe elaboration of adult bristles (Beckingham and Rubacha, 1984).The proposal that centriole migration and degradation is oneof the triggers for the switch from a mitotic to an endomitoticcycle is not a new one (Mahowald and Kambysellis, 1980), but theevidence is still somewhat scant. During stages 1-7 of oogenesis,112centrioles migrate from the nurse cells to the oocyte where theyare subsequently degraded. Although the reason for this migrationis not known, it does signal the onset of polyploidization of thenurse cells. Similarly, the migration and subsequent degradationof centrioles prior to polyploidization has also been described inthe follicle cells of developing ovarian cysts (Mahowald et al.,1979). Although centrioles have never been found in the terminallypolyploid cells of insects, there has been no direct demonstrationof a causal link between centriole behavior and entry into theendomitotic cycle.Although underreplication of the variegating locus has beeninvoked as one cause of PEV (Karpen and Spradling, 1990), there areseveral reasons why the su(var) phenotype associated with mfs(2)31mutations is probably not a consequence of ploidy of the affectedgene. First, a mutation which reduces the copy number of avariegating locus is more likely to exhibit an e(var) phenotypethan a su(var) phenotype. Second, and probably more important,mfs(2)31 alleles suppress PEV in the pigment cells of the eye, atissue which is generally considered to be diploid.Although a precise function of the mfs(2)31 gene product hasyet to be defined, it is hoped that additional functionalinformation can be obtained from a molecular analysis of the locus.Such an analysis is made possible by the isolation of mfs 2 , a Pelement induced allele of mfs(2)31. This allele reverts at afrequency of greater than 1% when passed through a dysgenicbackground and contains a P element insertion in subdivision 31E.113All of the phenotypes associated with mfs(2)31 mutations are co-reverted through mfs 2 indicating that they are all the consequenceof lesions in a single locus. Chapter 4 describes the use of thisallele to identify a putative transcript for the mfs(2)31 gene.114CHAPTER 4: MOLECULAR ANALYSIS OF THE mfs(2)31 LOCUS IntroductionThe mfs(2)31 locus is located in cytological subdivision 31Eon chromosome arm 2L of Drosophila melanogaster. Originallyisolated as a recessive sterile mutation with a tightly linkedbristle phenotype (Sandler, 1977), mfs(2)31 is a rare example of arecessive suppressor of position-effect variegation (su(var);Sinclair et al., 1991). We have had an ongoing interest in themfs(2)31 locus because of its su(var) phenotype.It has long been thought that the genes which can be mutatedto modify PEV will encode non-histone chromosomal proteins (NHPs;Spofford, 1976); a proposal which has been recently borne out bythe molecular analyses of several Su(var) loci. The protein whichis encoded by Su(var)205 (HP1) was purified from a nuclear proteinfraction and is known to associate with salivary gland polytenechromosomes (James and Elgin, 1986; James et al., 1989) while thepredicted Suvar(3)7 protein sequence contains widely spaced zinc-finger motifs (Reuter et al., 1990). It is becoming increasinglyapparent that PEV may be an effective tool for the geneticdissection of chromatin and/or the systems regulating its assembly.In this chapter we add mfs(2)31 to the relatively short list of PEVmodifiers which have been analyzed at the molecular level.Chapter 2 of this study describes the isolation of threeadditional alleles of mfs(2)31 including mfs 2 , a mutation which wasrecovered in a P element screen. The associated cytogeneticanalysis of the 31D-E region indicates that this locus is most115probably located in the distal half of 31E. Chapter 3 contains anextended phenotypic analysis of the mfs(2)31 locus including adescription of the new alleles. The mfs 2 allele is thought to behypomorphic, in that it is the only allele to survive in trans withthe original mfs 1 mutation. The mfs l/mfs 2 heteroallelic survivorsexhibit the same sterility, su(var), and bristle phenotypesobserved in mfs 1 homozygotes which strongly implies that all ofthese phenotypes are attributable to the mfs(2)31 locus.There are two reasons why we suspect that the mfs 2 mutation isthe result of an insertional event. First, the mfs2 straincontains a P element inserted into subdivision 31E which is themost probable cytological location of the mfs(2)31 locus (seechapter 3). Second, when passed through a dysgenic background, theallele reverts at a frequency of higher than 1% (see chapter 3)which is at least five orders of magnitude higher than spontaneousreversion frequencies for Drosophila melanogaster (Ashburner,1989). All the phenotypes associated with mfs(2)31 mutations,including the su(var) phenotype, are co-reverted through mfs 2 .This chapter describes the use of the P element containingmfs2 strain to identify a transcript which most likely correspondsto the mfs(2)31 locus. The locus has been molecularly mapped withrespect to the previously cloned and tightly linked da and cdc2genes and a transcriptional profile of the region is described. Afull length cDNA corresponding to the mfs(2)31 transcript has beenisolated and sequenced. The similarity of the deduced gene productto previously described proteins, and the implications of these116similarities with respect to the usefulness of PEV as a geneticassay, are discussed.117Materials and MethodsSTOCKSComplete descriptions of all strains which have been used forRNA and DNA isolations can be found in chapter 3 and/or in Lindsleyand Zimm (1991).RECOMBINANT CLONESThe pn25.1 construct is a full length P element, plus flankinggenomic DNA, cloned into pBR322 (described in O'Hare and Rubin,1983). The pDmA2 construct contains a full length Drosophila actingene plus flanking genomic DNA cloned into pBR322 (described inFryberg et al., 1980). JT20, JT31 and JT35 are overlapping clonesfrom subdivision 31E inserted into a modified CoSpeR vector. Thesecosmids, all of which encompass the da locus, were kindly providedby C. Cronmiller. The host bacterium for each of these recombinantclones is Escherichia coli (E.Coli) DH5a.IN SITU HYBRIDIZATIONSIn situ hybridizations to polytene chromosomes were performedas previously described (see Chapter 3: Materials and Methods).DNA ANALYSISIsolation of high-molecular weight genomic DNA Genomic DNA was isolated using a modified procedure of Jowett(1988). Approximately 1-3 g of tissue were flash frozen in liquid118nitrogen and ground to a fine powder in a pre-chilled mortar andpestle. The powdered tissue was immediately transferred to 10 mlof lysis buffer (10 mM Tris-HC1 (pH 8.0), 50 mM NaC1, 50 mM EDTA,1% SDS, 0.15 mM spermine, 0.5 mM spermidine) and Proteinase K(stock solution is 10 mg/ml in 50% glycerol) was added to a finalconcentration of 100 ug/ml. Following a two hour incubation at37°, the mix was extracted twice with phenol ( BRL: equilibratedwith TE (pH 8.0)), twice with phenol/chloroform (1:1 v/v), and oncewith chloroform. The aqueous solution was overlayed with 5 ml ofice-cold 95% EtOH and the DNA was spooled out using a narrow boreglass Pasteur pipet. The DNA was suspended in 500 ul of TE (pH8.0) and RNase (Sigma) was added to a final concentration of 100ug/ml. Following a 60 minute incubation at 37°, the reaction wasextracted once with phenol (as above), once with phenol/chloroform(1:1 v/v) and once with chloroform. The aqueous phase wasoverlayed with 500 ul of cold 95% EtOH and the DNA was spooled asabove. Depending on yield, the DNA was suspended in 200-1000 ul ofTE (pH 8.0) and stored at 4°.Isolation of plasmid and cosmid DNA Procedures for purifying either plasmid or cosmid DNA wereidentical in all respects. All protocols utilized modifications ofthe alkaline-lysis method of DNA isolation (Sambrook et al., 1989).Small scale isolations: Small scale DNA isolations wereobtained using the Magic Miniprep DNA purification system119(Promega). Overnight cultures (1-3 ml of LM; 0.005% ampicillin)were pelleted, suspended in 200 ul of Resuspension Solution (50 mMTris-HC1 (pH 7.5), 10 mM EDTA, 100 ug/ml RNase A) and then lysedwith 200 ul of Cell Lysis Solution (0.2 M NaOH, 1% SDS). Bacterialdebris was precipitated with 200 ul of Neutralization Solution(2.55 M KOAc (pH 4.8)) and removed by centrifugation at 15,000 rpmfor 5 minutes. One ml of resuspended Purification Resin was mixedinto the supernatant and the slurry was forced (by syringe) througha mini-column attached to a 3 ml disposable syringe barrel.Columns were washed with 3 mis of Column Wash Solution (200 mMNaC1, 20 mM Tris-HC1 (pH 7.5), 5 mM EDTA; diluted 50% with absoluteethanol) and dried by centrifugation. DNA was eluted in 50 ul ofpre-heated (65°) TE buffer (pH 7.5).Medium scale isolations: Overnight cultures (100 ml LM;0.005% ampicillin) were pelleted and then suspended in 2.5 ml ofSolution 1 (50 mM Glucose, 25 mM Tris (pH 8.0), 10 mM EDTA).Following a ten minute incubation on ice, the cells were lysed with5 ml of freshly prepared Solution 2 (0.2 M NaOH, 1% SDS) andprecipitated with 3.75 ml of 3 M KOAc. The bacterial debris waspelleted by centrifugation at 8000 rpm for 20 min in a Sorvall ss34rotor. The clear supernatant was extracted once withphenol/chloroform (1:1) and once with chloroform prior toprecipitation of the DNA with two volumes of 95% EtOH. The DNA waspelleted, washed with 70% ethanol and vacuum dried. The pellet wassuspended in 300 ul of TE and RNase (Sigma) was added to a finalconcentration of 100 ug/ml. Following a 30 minute reaction time,120the RNase was precipitated with 1/2 volume 5M NH 4Ac (-70° for 20minutes) and removed by centrifugation at 15,000 rpm for 5 min.The DNA was precipitated with two volumes of 95% EtOH, pelleted,vacuum dried and suspended in 900 ul of TE (pH 7.5). Plasmid DNAwas precipitated (one hour on ice) with 600 ul of PEG solution (20%PEG, 2.5 M NaC1). The DNA was pelleted (15,000 rpm for 10 min),washed in 70% EtOH, vacuum-dried and suspended in 100 ul of TE.Large scale isolations: Overnight cultures (500 ml of LM;0.005% ampicillin) were pelleted and then suspended in 4 ml ofSolution 1 (50 mM glucose, 25 mM Tris (pH 8.0), 10 mM EDTA).Following a ten minute incubation on ice, the cells were lysed with16 ml of freshly prepared Solution 2 (0.2 N NaOH, 1% SDS). Thebacterial debris was precipitated with 12 ml of 3 M KOAc andremoved by centrifugation (20 min at 15,000 rpm in a Sorvall ss34rotor). The DNA was precipitated in isoproponal (60% v/v),pelleted, washed in 70% EtOH, vacuum dried and suspended in 8 ml ofTE (pH 8.0). The suspended DNA was prepared for centrifugation bythe addition of eight grams of solid cesium chloride and 650 ul ofethidium bromide (10 mg/ml in H 20). The solution was transferredto a Beckmann polyallomer 16 X 76 Quick-Seal centrifuge tube andthe tube was heat-sealed. Centrifugation was at 45,000 rpm for 36hours (20°) in a Sorvall Tft65.13 rotor in a Beckmann L8-80ultracentrifuge. Supercoiled DNA was removed from the cesiumchloride gradient using a 21 gauge needle attached to a 1 mlsyringe barrel. The ethidium bromide was removed by repeatedextractions with water saturated 1-butanol. The solution was121diluted three times with water and the DNA was precipitated with2.5 volumes of 95% EtOH.Isolation of bacteriophage DNA Confluent plates of bacteriophage X were prepared as follows.Approximately 3 X 10 6 bacteriophage were mixed with 1.6 X 10'bacterial cells and then incubated at 37° for 30 minutes. The mixwas then diluted in 8 ml of melted top agar (cooled to 43°) andimmediately spread on the surface of LM plates. The plates wereincubated at 37° for 8 hours or until complete plate lysisoccurred. The top agar was scraped from the surface of the platesinto a solution of 25 ml X-dil: 1 ml CHC1 3 and then left overnightat 4°. The debris was removed by centrifugation for 30 min at 2500rpm in a desktop clinical centrifuge. The supernatant wastransferred to fresh tubes and made 1 M NaC1 (2.92 g/100 mlsupernatant) and 10% polyethylene glycol (10 g/100 ml supernatant).The mixture was incubated for one hour in ice water to allowprecipitation of the bacteriophage particles which were thenpelleted by centrifugation at 8000 rpm for ten minutes (4°) andsuspended in 0.5 ml of X-dil.Cesium chloride step gradients were prepared in the followingmanner. One ml of CsC1 Solution 1 (5.0 m CsCl, 10 mM Mg50 4 , 10 mMTris (pH 8.0), 0.1 mM EDTA) was placed in the bottom of a Beckman13 X 51 mm Ultra-Clear centrifuge tube. This was overlayed with 3ml of CsC1 Solution 2 (3.0 M CsCl, 10 mM MgSO 4 , 10 mM Tris (pH8.0), 0.1 mM EDTA) which in turn was overlayed with the 0.5 ml122bacteriophage solution. Centrifugation was in a Beckman SW50.1rotor at 30,000 rpm for 1 hour (20°). Bacteriophage were removedfrom the gradient with a 21 gauge needle attached to a 1 ml syringebarrel. One volume of 1 M Tris (pH 8.0), 0.25 volumes of EDTA and2.25 volumes of formamide were added directly to the solution.Following a 15 minute incubation at room temperature, the DNA wasprecipitated with 2.5 volumes of 95% EtOH, pelleted, washed in 70%EtOH and resuspended in 100 ul of TE.Restriction endonuclease digests Approximately 1-3 ug of DNA (plasmid, cosmid or genomic) wasdiluted in 16 ul of water. Two ul of the appropriate 10 X corebuffer (BRL) and two ul of restriction enzyme (3-10 units) wereadded and the reaction was incubated at 37° for a minimum of onehour. Reactions were stopped by the addition of 5 ul of loadingbuffer (6 M Urea, 25% Sucrose, 50 mM EDTA, 0.4% xylene cyanol, 0.4%bromophenol blue).Gel Electrophoresis DNA was size fractionated on 0.6-1.2% agarose (BRLelectrophoresis grade) gels. The gel buffer contained 40 mM Tris-HC1, 5 mM NaOAc, 1mM EDTA; solution is pH adjusted to 7.8, and 0.1ug/ml ethidium bromide. Electrophoresis was at 1.5-4 v/cm for 1-18hours. The gels were photographed using transmitted ultravioletlight for illumination.123DNA transfer from agarose gels to nylon membrane DNA was transferred to nylon membranes using a modifiedprocedure of Southern (1975). Gels were gently agitated indenaturing solution (0.4 N NaOH, 0.6 M NaC1) for 30 min at roomtemperature, rinsed briefly in H 2O and then neutralized in 1.5 MNaC1, 0.5 M Tris-HC1 (pH 5.5), for an additional 30 minutes. Thegel was then placed on a pre-soaked (20 X SSC) Whatman 3mm paperwick, which was suspended on a platform above a reservoir filledwith 20 X SSC. The gel was overlayed first with a pre-wetted nylonmembrane (as per manufacturers instructions), then two thicknessesof pre-soaked (20 X SSC) Whatman 3 mm paper and finally, a thickstack of paper towels. A weighted plexiglass plate was placed ontop to ensure even transfer. Transfer was for 3-12 hours forplasmid and cosmid DNA and 36 hours for genomic DNA. The papertowels were replaced periodically as required. Following transfer,filters were air-dried for two hours and then the DNA was fixed tothe membrane by exposure to ultraviolet light (254 nm) for fiveminutes.RNA ANALYSISIsolation of developmentally staged Drosophila cultures In order to observe the transcriptional profile of mRNA's atdifferent times in development, the following staged cultures werepurified from the Oregon-R wild-type strain; 0-2.5 hour embryos, 0-6 hour embryos, 2.5-6 hour embryos, 6-9 hour embryos, 9-12 hourembryos, 12-24 hour embryos, 1st instar larvae, 2nd instar larvae,1243rd instar larvae, early pupae, late pupae and adult male andfemales. In addition, 0-6 hour embryo's were isolated for the Infs l ,Infs 2 , mfs3 , mfs 4, b pr cn, mfe2, mfsR-3, mfsR4, and mfsR5strains. Population cages were established for each strain at 25°.Egg lays were on plates made up of 1% agar, 5% malt vinegar and 5%EtOH. A live yeast paste (Fleischmann's) was spread on the surfaceof the plates to encourage oviposition. Synchronized cultures wereobtained in the following manner.Embryonic collections:  The duration of an egg-laycorresponded to the length of the embryonic period for which tissuewas required i.e. 3 hours for 3-6 hour embryos, 12 hours for 12-24hour embryos. Embryos were aged at 25° for the appropriate periodof time and were then collected and concentrated by gentlemanipulation on a fine nylon mesh. Embryos were transferred tomicrocentrifuge tubes, weighed and flash frozen by immersion inliquid nitrogen. Embryos were stored at -70° until needed.Larval and pupal collections: Egg-lays for larval and pupalcollections were for 24 hours on standard cornmeal-agar fly mediain the bottom of 6 X 12" Tupperware containers. Subsequent to egg-lays, the containers were removed from the population cages, sealedand stored at 25°. First instar larvae were gently brushed fromthe surface of the food while second and third instars wereseparated from the food by floatation in 10% sucrose. Appropriatelystaged pupae were selected visually.125Isolation of Poly(A)+ RNA To ensure inactivation of contaminating nucleases and preventsubsequent contamination, all solutions were treated with 0.1%Diethylpyrocarbonate (DEP) where possible. In addition, allglassware was washed using DEP treated water and baked overnight at250°.Whole RNA was extracted using a modified procedure of Jowett(1988). Approximately 200-1000 mg of pre-frozen tissue wereimmersed in liquid nitrogen and ground to a fine powder using amortar and pestle. The powder was immediately transferred to 4 mlof pre-heated phenol (65°) saturated with 0.2 M NaOAc (pH 5.0).The solution was vortexed and then 3 ml of NaOAc and 0.8 ml of 10%SDS was added. Following a 5 minute incubation at 65°, thesolution was again vortexed and allowed to cool to roomtemperature. Four mis of chloroform were added, the solution wasvortexed again, and the aqueous phase was separated bycentrifugation. The aqueous phase was extracted once withphenol/chloroform (1:1 v/v) and once with chloroform alone. TheRNA was precipitated with 2.5 volumes of 95% EtOH and stored at-20° under ethanol.Poly(A)+ RNA was isolated from the whole RNA by chromatographyon Oligo(dT)cellulose columns. Oligo(dT)cellulose (Pharmacia),which had been equilibrated in 1 X Binding Buffer (10 mM Tris (pH7.5), 0.4 M NaCl, 5 mM EDTA, 0.5% SDS), was used to prepare 0.5 mlcolumns in 12 ml Bio-rad econo-columns. The prepared columns werewashed first with 5 volumes of 0.1 M NaOH, then 10 volumes of126sterile water and finally 10 volumes of 1 X Binding Buffer.RNA (stored under ethanol) was pelleted, resuspended in 10 mlH 2O (DEP) and then diluted in an equal volume of 2 X BindingBuffer. The RNA solution was heated to 65° for five minutes andthen loaded on the column. The column was then washed, first with10 volumes of Binding Buffer and then with an additional 10 volumesof Binding Buffer containing no SDS. The RNA was eluted in 3 X 1ml fractions with Elution Buffer (10mM Tris (pH 7.5), 1 mM EDTA).The eluted RNA solution was made 0.3 M NaOAc and the RNA wasprecipitated with 2.5 volumes of 95% EtOH. The poly(A)+ RNA wasstored at -20° as a precipitate in ethanol.Gel electrophoresis of RNA Poly(A)+ RNA was size fractionated on formaldehyde (0.66 M)denaturing gels. Two grams of agarose were dissolved in a solutionof 175 ml DEP treated water and 20 ml 10 X MOPS Buffer (0.2 M MOPS,50 mM NaOAc, 10mM EDTA; pH to 7.0). The solution was cooled to 50°prior to adding 10.2 ml of 37% de-ionized formaldehyde. The gelwas poured and allowed to set for 1 hour before use. The wellswere flushed prior to the loading of samples.Approximately 5 ug of RNA (stored as a precipitate in ethanol)was pelleted and resuspended in 11 ul water, 2 ul 10 X MOPS Buffer,7 ul formaldehyde, 20 ul formamide and 2 ul 1 mg/ml Ethidiumbromide. The sample was heated to 55° for five minutes, 2 ul ofsterile loading buffer (6 M urea, 25% sucrose, 50 mM EDTA, 0.4%xylene cyanol, 0.4% bromophenol blue) were added and then the127sample was loaded onto the gel. Electrophoresis was at 60 V for 8hours. The gels were photographed using transmitted ultravioletlight.RNA transfer from formaldehyde gels to nylon membranes The formaldehyde gels were prepared for transfer by soakingfor 2 X 20 minutes in 10 X SSC. In all other respects transferswere performed as described above for agarose gels (see DNAtransfer from agarose gels to nylon membrane).HYBRIDIZATION TO NYLON MEMBRANESIn all respects, hybridization to DNA and RNA filters wasconducted in the same manner.Gel purification of probes To prepare probes for radiolabelling, plasmid or bacteriophageDNA was cut with the appropriate restriction endonuclease and size-fractionated on agarose gels. A small gel slice containing thefragment of interest was excised from the gel and transferred to amicrocentrifuge tube. The DNA was extracted from the agarose byadhesion to glass beads (GENECLEAN; Bio 101) as per manufacturersinstructions.Radiolabellinq of probes DNA was radiolabelled to a specific activity of approximately1.8 X 10 9 dpm/ug by random hexamer-primed DNA labelling (Feinberg128and Vogelstein, 1983) using the Boehringer Mannheim Random PrimedDNA Labelling Kit. Radiolabelled fragments were purified onSephadex columns (G50 DNA grade; Pharmacia) which were prepared asfollows. A suspension of Sephadex in sterilized TE (pH 7.5) wasused to build six inch columns in modified Polyethylene TransferPipets (Fisher Scientific). Columns were allowed to pack bygravity for 30 minutes. Radiolabelling reactions were stopped byaddition of an equal volume of Column Loading Buffer (40 mM EDTA,2% Blue Dextran). The reaction was loaded on the column and theblue fraction was collected. The radiolabelled fragment was boiledfor five minutes and plunged into iced water prior to adding tohybridization bags (see below).Filter Hybridization Filters were sealed in plastic bags with 10 ml pre-hybridization buffer (5 X SSC, 5 X Denhardt's, 0.5% SDS, 100 ug/mlherring sperm DNA) and incubated at 65° for one hour with gentleagitation in a shaking water bath. A prepared radiolabelledfragment (see above) was added directly to the bag. Hybridizationwas at 65° for 12-18 hours. High stringency washes were performedas follows: 2 X 1 hour in 2 X SSC, 0.1% SDS (65°)followed by 2 X 1hour in 0.1 X SSC, 0.1% SDS (65°). Hybridization was detected byautoradiography.129LIBRARY CONSTRUCTION AND SCREENINGConstruction of a bacteriophage library To facilitate cloning of the mfs(2)31 gene, a bacteriophage Xlibrary was prepared from mfs 2 DNA. Two micrograms of EMBL3 vectorDNA (Promega) were digested to completion with Eco R1 (10 units).The digested DNA was diluted to 100 ul with H 2O, made 0.3 M NaOAc,and precipitated with 2.5 volumes 95% EtOH.Ten micrograms (ug) of high-molecular weight genomic DNA fromthe mfs2 strain was digested at 37° with 0.1 units of Eco R1. Thetime course of the reaction was monitored by removing 1 ug aliquotsof DNA. The aliquots were inactivated immediately by making them20 mM with respect to EDTA. One hundred ng of DNA from eachaliquot was examined by electrophoresis on an agarose gel todetermine which sample had the greatest quantity of DNA falling inthe 15-20 kb range. This sample was retained while the others werediscarded.The partially digested genomic DNA was precipitated in 2.5volumes 95% EtOH, washed with 70% EtOH and suspended in phosphatasebuffer (50 mM Tris (pH 9.0), 1 mM MgC1 2 , 0.1 mM ZnC1 2 , 1 mMspermidine). One unit of calf alkaline phosphatase was added andthe reaction was incubated at 37° for one hour. The reaction mixwas extracted once with phenol/chloroform (1:1 v/v) and once withchloroform. The sample was made 0.3 M NaOAc and precipitated with2.5 volumes of 95% EtOH.Two ug of vector DNA and 0.5 ug of the dephosphorylatedgenomic DNA were co-precipitated and resuspended in 10 ul ligase130buffer (10 mM Tris (pH 7.5), 10 mM MgC1 2 , 5 mM (3-mercaptoethanol,1 mM ATP). One unit of T4 DNA Ligase (BRL) was added and thereaction was incubated overnight at 16°. The ligated DNA waspackaged using the Gigapack Plus (Stratagene) packaging kitaccording to the manufacturers instructions.Screening Bacteriophage libraries Two bacteriophage X libraries were screened in this study.The first was an EMBL3 genomic library constructed from partiallydigested (Eco R1) mfs 2 DNA (see above). The host strain bacteriumfor this library is NM539 (E. coli). The second library screenedwas an imaginal disc cDNA library (Dr. G. Rubin) constructed inXgt10. This library was amplified on C600hf1A (E. coli). Bothlibraries were screened using a modified procedure of Benton andDavis (1977).Preparation of confluent plates: A single colony of theappropriate host strain was grown overnight at 37° in 50 mis of LM.Five mis were used to inoculate a fresh 50 ml culture and was grownat 37° for 3-5 hours. The cells were pelleted at low speed in adesk-top clinical centrifuge, suspended in 25 ml 10 mM MgSO 4 andstored at 4° for up to 4 weeks.Approximately 5 X 10 4 bacteriophage were added to 100 ul ofthe host cells, and were then incubated at 37° for 30 minutes. Themix was then added to 8 ml of molten top agar which had been cooledto 53°. The top agar was immediately poured onto the surface of LMplates which were then incubated at 37° until lysed bacterial lawns131covered the plates.Plague lifts: Lysed plates were carefully overlayed with dry,pre-cut nitrocellulose membranes. The membranes were peeled offthe plates, soaked in 0.4 N NaOH for 30 seconds, washed twice for60 seconds in 1.5 M NaC1, 0.5 M Tris (pH 5.5) and then washed for30 seconds in 2 X SSC. The filters were air-dried and baked fortwo hours at 80°. Duplicate filters were made for each plate.Hybridization of the filters to radiolabelled probes was conductedas previously described (see Filter hybridization). Following thedetection of homology, the appropriate agarose plug was recoveredfrom the plate and successive rounds of screening were conducted athigher bacteriophage dilutions until plaque-purity was achieved.Screening cosmid libraries A cosPneo library prepared from Drosophila DNA wasgenerously provided by J. Leung. The DNA was prepared by partialMbo 1 digestion (35-50 kb size range) and the library was amplifiedin the recA E. coli strain Dill (Leung, 1988). The library wasplated overnight on LM agar plates (0.005% ampicillin) and the DNAwas blotted directly onto the surface of nitrocellulose filtersusing the following modified protocol of Grunstein and Hogness(1975).Pre-cut nitrocellulose filters were overlayed on the surfaceof the plates, peeled-off and placed immediately (bacterial sideup) on Whatman 3MM paper soaked with denaturing solution (0.5 MNaOH, 1.5M NaCl). Following 5 minutes on the denaturing solution,132the filters were transferred to 3MM paper soaked in neutralizingsolution (1.5 M NaC1, 0.5 M Tris (pH 8.0)) for an additional 5minutes. Filters were air-dried and baked at 80° for 2 hours.Hybridization of the filters to radiolabelled probes was aspreviously described (see Filter hybridization).Subcloninq cosmid, plasmid and bacteriophaqe DNAFragments of interest were cut from the vector and separatedon an agarose gel. The appropriate band was excised from the geland purified using the Geneclean DNA purification kit (Bio 101)according to the manufacturers instructions. Approximately 50-200ng of DNA was co-precipitated with 50 ng of pre-cut pUC 19 vectorwith compatible ends. The DNA was resuspended in 20 ul of 10 mMTris (pH 8.0), 10 mM MgC1 2 , 1 mM DTT and 1 mM ATP. One unit of T4DNA ligase was added and the reaction was incubated overnight at16°. Following incubation, 5 ul of the ligation mix was added to100 ul of sub-cloning efficiency E. col]: DH5a competent cells(BRL), which were then incubated for a further 30 minutes on ice.The cells were then heat shocked for 45 seconds at 42°, andreturned to the ice for another 2 minutes. The mix was dilutedwith 1 ml of LM and incubated in a rotary shaker for 1 hour at 37°.Approximately 50-200 ul of cells were spread evenly on LM agaroseplates (0.005% ampicillin) whose surfaces had been pre-treated with40 ul of 3% X-gal solution. Recombinants were identified as whitecolonies and screened by mini-plasmid preparations.133DNA SEQUENCINGSequencing reactions All sequencing reactions were double-stranded using Puc 19 asthe template vector. Four ug of plasmid DNA was denatured bysuspending it in 20 ul water and 2 ul of a 2 M NaOH, 2 mM EDTAsolution, and then heating the solution to 65° for 5 minutes.Following denaturation, the DNA was precipitated with 2 ul 2 MNH 4Ac and 55 ul 95% EtOH for 30 minutes at -20°. The denaturedtemplate was pelleted, washed in 70% EtOH, vacuum-dried andsuspended in 7 ul water. One ul of each primer (3 pmoles) and 2 ulAnnealing Buffer (50 mM NaCl, 10 mM MgC1 2 , 40 mM Tris (pH 7.5))were added to the denatured template DNA. Annealing was at 37° for20 minutes followed by incubation at room temperature for anadditional 10 minutes. The labelling reaction was preparedsequentially by combining in the following order: 10 ul annealedtemplate, 1 ul 0.1 M DTT, 2 ul nucleotide mix (1.5 uM each of dCTP,dGTP and dTTp), lul (a- 35S)dATP and 2 ul (1.8 Units) T7 DNApolymerase. Following the addition of the T7 polymerase, thereaction was incubated at room temperature for seven minutes.Teillination reactions were performed by adding 3.5 ul of thelabelling reaction to four tubes, each of which contained 2.5 ul ofa different termination mix. A termination mix consists of 80 uMeach of 3 of the 4 dNTPs found in DNA and 8 ul of the fourth. Thetermination reactions were incubated at 37° for 10 minutes and thenall enzymatic activity was stopped by the addition of 4 ul of StopSolution (95% formamide, 20 mM EDTA, 0.5% bromophenol blue, 0.5%134xylene cyanol).Polyacrylamide gel electrophoresis Fifty ml of gel solution was prepared by combining 23 g urea(final molarity is 7 M), 13.2 ml 5 X TBE (Sambrook, 1989), 5.5 mlLong Ranger Gel Solution (J. T. Baker) and bringing to volume withwater. Polymerization was induced by the addition of 250 ul 10%ammonium persulphate and 25 ul TEMED.Plates were cleaned of any debris with 95% EtOH. One platewas treated with Dimethylchlorosilane solution (BDH) and re-cleanedwith ethanol. Plates (33 X 40 X 0.04 mm) were assembled accordingto manufacturers instructions. Sequencing reactions were loadedinto sharktooth sample wells and separated by electrophoresis at 45watts for 2-6 hours. Gels were dried for two hours prior toautoradiography.Seguencing strategy All sequencing reactions were double-stranded using pUC19 asthe sequencing vector. Following initial sequencing reactions,using the universal forward and reverse primers, primers weresynthesized for each successive reaction based on the sequenceobtained from the previous reaction. Primers were spacedapproximately 150-200 by apart and covered both strands. Both thecDNAs and the corresponding genomic fragments were sequenced onboth strands using the same set of primers.Oligodeoxynucleotide primers used for sequencing were135synthesized by the UBC Oligodeoxynucleotide Synthesizing Facility.Deprotected primers were purified as described by Sawadogo and VanDyke (1991) : 100 ul of primer solution in 30% NH 4OH was vortexedvigorously in a 1.5 ml eppendorf tube with 1000 ul n-butanol andthen centrifuged for 1 min at 12,000 rpm. The single n-butanolphase was discarded and the oligonucleotide pellet was dissolved in100 ul water.136ResultsCloning of a P element from mfs 2 To determine the number of P elements in the mfs2 b pr cn/CyOstrain, high molecular weight genomic DNA was digested with Bam HIand hybridized with a 900 by Hind III fragment cut from the Pelement in pn25.1 (Figure 10). Bam HI does not cut within thecomplete P element and thus, is a useful restriction enzyme fordetermining the total number of P elements present in a genome.Two hybridization signals are detected in the mfs 2 DNA which arenot present in the control lane: a strong signal at 1.2 kb and aweak signal at 9.0 kb (Figure 10). Although, several repeats ofthis experiment have failed to detect any additional elements, theweak hybridization at 9.0 kb is reproducible.Since the mfs2 b pr cn/CyO strain contains only two detectableP element homologies, it was not necessary to remove elements fromthis stock by recombination. An EMBL 3 genomic library, wasconstructed from partially digested (Eco RI) mfs 2 b pr cn/CyO DNAand was screened with the 900 by Hind III P element probe describedabove. Out of 250,000 plaques screened, four positives were andplaque-purified through successive rounds of hybridization. DNAfrom the four X clones was digested with Eco RI (Figure 11a), andhybridized to the 900 by P element fragment used to screen thelibrary (Figure 11b). Three of the four clones (X2, X3 and X4)have similar Eco RI restriction digest patterns and contain a 1.5kb fragment which hybridizes strongly to the P element probe. Thefourth clone (X1), has a different Eco RI restriction digestion137Figure 10: mfs 2 b pr cn/CyO genomic DNA hybridized with the 900 byHind III fragment from pn25.1.Lane 1^X Hind III2 mfs2 b pr cn/CyO^Hind III1^28.5 kb1.5 kb138Figure 11: Bacteriophage X clones hybridized with the 900 by HindIII fragment from pn25.1.Lane^1^X^Hin dIII2^X1 Eco RI3^X2^Eco R14 X3 Eco R15 X4^Eco R1A. Photograph of ethidium bromide stained gel.B. Photograph of autoradiogram.A^ B1 2 3 4 5^ 1^2^3^4^58.5 kb^IMO1.5 kb139pattern and contains a weak P element homology in an 8.5 kbfragment.In situ hybridizations were done to Oregon-R salivary glandpolytene chromosomes using probes made from the full length X1 andX2 clones. The X2 probe hybridizes to the base of chromosome arm2R (data not shown), while the X1 probe hybridizes to subdivision31E on chromosome arm 2L (Figure 12), the known cytologicallocation of the mfs(2)31 locus. Since the X1 probe does nothybridize to any other site in the Drosophila genome, I presumethat the X1 clone contains the P element which is inserted into 31Ein the mfs 2 strain (see Chapter 3).Mapping Xl. To facilitate the mapping of the P element in 31E relative totranscripts in the region, a restriction map was generated of Xl.Overlapping fragments from X1 were subcloned into pUC19 and mappedusing a battery of seven restriction enzymes: Pst I, Xba I, Sal I,Eco RI, Sst I, Hind III and Bam HI. Figure 13 describes thesubcloned fragments from Xi while Figure 14 contains the compositerestriction map generated from these constructs. Regions ofoverlap between subclones and any mapping ambiguities have beenresolved and/or verified by hybridization.To determine the orientation of the P element with respect tothe previously cloned da and cdc2 genes, subcloned fragments fromX1 were hybridized with probes made from the full length JT31 andJT35 constructs (Figure 15). JT31 and JT35 are overlapping140Figure 12: Localization of X1 to subdivision 31E by in situhybridization.141Figure 13: Restriction maps for various fragments from lambda 1. Allfragments have been cloned into Puc 19: R = Eco R1, S =Sail, B = Bam HI, T = Sstl, P = Pst 1, X = Xba I, H =Hind Ill.RS B^ RpR8.5^1 1^I 1^I I I^I^1^IHTP P^X H R R SRpP6.0^i^I^I^I^I I^I I^1P X H H PpS5.0SR^R BBR^B S1 I I^i ■ j I^II^1^I^IP H H XpB2.8B S^R^R Bl^1 i 1^I1 1^1P^X PS^R^R SpS2.4^11^1^iI i^IP X PpB2.2^7 1^11 H X^BB R RS BpB1.5^I (II ^ ^HT TPpH1.5H^RH^H 1 kbpH1.0142Figure 14: Composite restriction map for Lambda 1. The lines at the top of the figure indicate thealignment of the various subcloned fragments. The thick bar indicates the region of Pelement homology: R = Eco RI, S = Sal I, B = Barn HI, T = Sst I, P = Pst I, X = Xba I, H = Hind III.i--4.(0^ pS5.0 ^ pP6.0 ^pH1 .5pH1 .0pB2.8pB2.2 ^pB1 .5 pS2.4R R RS B^ RR SR^R BBR^BS^R^R RL (II ^1 i 1^1 i^I^I II I^I^I^[^1I^II I^I^I^I^I 1^1^1^I I^IT TP P X^H^H P^H H X^P^X PpR8.51 kbFigure 15: Overlapping subcloned fragments from X1 hybridized withB. JT31 and C. JT35.Lane^1^X^Hind III2^pB1.5^Bam HI3^pR8.5 Eco RI4^pP6.0^Pst I5^pS5.0 Sa1 I6^pB2.8^Bam H1B (JT31)1^2^3^4^5^6144cosmids, which both encompass the da and cdc2 genes: JT31 extends10 kb more proximally while JT35 extends 15 kb more distally(Clegg, 1992). The orientation of the two cosmids with respect tothe centromere is inferred from the mapping experiments describedin Chapter 2 which place the da gene proximal to the cdc2 gene.The plasmids pB2.9, pS5.0, pP6.0, pR8.5 and pB1.5 containoverlapping fragments which completely span X1 (restriction mapsare provided in Figure 13). While there is no homology betweenJT35 and these Al fragments (Figure 15b), JT31 does overlap,extending into the 6.0 kb Pst 1 fragment (pP6.0) which contains theP element homology (Figure 15c). This would indicate that themolecular order of the three genes is, moving distal to proximal,cdc2-da-mfs(2)31 (Figure 16), thus confirming the recombinationmapping in chapter 2.The P element homology in X1 is completely contained within an8.5 kb Eco RI (pR8.5) fragment. To further localize the P elementsequences, this fragment was digested with Hind III and hybridizedto the full length pn25.1 plasmid (Figure 17). The P elementhomology in pR8.5 is contained within a 1.0 kb Hind III fragment.The position of this fragment is indicated in Figure 14. Since afull length P element is 2.9 kb in length, the element in 31E mustbe degenerate.To determine whether the P element homology in X1 correspondsto a mobilizable element, high molecular weight genomic DNA wasextracted from the following strains; b pr cn/CyO, mfs 2 b prcn/CyO, mf -Ris b pr cn/CyO, mfe2 b pr cn/CyO, mfsR3 b pr cn/CyO, mfs R4145Figure 16: Overlapping clones in subdivision 31 E. The precise extent of overlap between JT31 andLambda 1 has not been determined and thus is indicated with a dashed line. The locationof cloned genes within the constructs is indicated. The orientation of the clones withrespect to the centromere is inferred from the recombination mapping described in chapter2.A.H,^cdc2 da^ mfs(2)31JT35 ^JT31 ^^Lambda I ^Centromereb pr cn/CyO and mfsR5 b pr cn/CyO. The mfel , mfsR2 , mfaR3 , mfs R4 andmfsR5 strains are hybrid dysgenesis induced revertants of the mfs 2mutation (see Chapter 3). Genomic DNA from each of these strainswas doubly digested with Eco RI and Xba I and hybridized with the4.0 kb Eco RI/Xba I fragment cut from pR8.5 (Figure 18). Thisfragment spans the region of P element homology in Xi (see Figure14). While this probe hybridizes to a single 4.0 kb band in the bpr cn/CyO control DNA (Lane 1), it hybridizes with equal intensityto bands of 4.0 kb and 4.5 kb in the mfs 2 b pr cn/CyO DNA (Lane 2).This polymorphism could indicate the loss of either the Eco RI orXba I sites on the mfs2 chromosome or alternatively, it couldindicate the presence of a 500 by duplication or insertion. Sincethe polymorphism is lost in all five revertant strains (Lanes 3 - 7),I feel that this result indicates the presence of a 500 bymobilizable P element contained within Xl.Transcriptional analysis of X1 and the mfs(2)31 locus. To identify transcripts which the inserted P element may beinterfering with, all the embryonic messages encompassed by X1 weremapped. Poly(A)+ RNA was isolated from 0-6 hr wild-type embryosand probed with contiguous fragments from Xl. The results ofresults of the transcriptional analysis are summarized in Figure19. In cases where hybridization failed to detect any message, thefilters were re-probed with pDmA2 (actin) to ensure that thetransfer of the RNA was successful.As indicated, a large number of embryonic transcripts were148Figure 17: Digested 8.5 kb Eco RI fragment from X1 hybridized withpn25.1.Lane^1^X^Hind III2^pR8.5^Eco RI3^pR8.5 Eco RI/Hind III4^pR8.5^Hind IIIA. Photograph of ethidium bromide stained gel.B. Photograph of autoradiogram.A^ B1^2^3^4 1^2^3^4IMO11111111 ado 014110 1.0 kb147Figure 18: mfs(2)31 b pr cn/CyO genomic DNA hybridized with a 4.0kb Eco RI/Xba I fragment from X1.Lane 1 b pr cn/Cy0 Eco RI/Xba I2 mfs2 b pr cn/Cy0 Eco RI/Xba I3 mfsR1 b pr cn/Cy0 Eco RI/Xba I4 mfsR2 b pr cn/Cy0 Eco RI/Xba I5 mfsR2 b pr cn/Cy0 Eco RI/Xba I6 mfsm b pr cn/Cy0 Eco RI/Xba I7 mfsR5 b pr cn/CyO Eco RI/Xba I1^2^3^4^5^6^74.5 kb4.0 kb 01.111401011141100M11,149Figure 19: Transcriptional analysis of Lambda 1. The solid lines at the top of the figure represent theprobes which were used to detect the messages indicated at the bottom of the figure. Thesizes of the messages are indicated. The thick bar indicates the region of P elementhomology. Abbreviations are the same as in Figure 14.R R RS B^ RR SR^R BBR^BS^R^R RL III^I I^I^I I^I^I I J I^I^I^I^Iill I^i^I^I^ I i^I^I^I I^IT TP P^X^H^H P^H H^X^P^X P4. 1 kbidentified in this analysis. Of particular interest is the regionin the immediate vicinity of the P element insertion where nomessages are detected within 7 kb of the distal end of the Pelement containing fragment. The 1.5 kb Hind III/Eco RI fragmentimmediately proximal to the element detects 0.8 and 1.2 kbtranscripts while the next most proximal fragment detects a 0.8 kbtranscript only. It is possible that the 0.8 and 1.2 kb messagesdetected by the 1.5 kb Hind III/Eco RI fragment are differentiallyprocessed products of the same transcription unit while the 0.8 kbtranscript detected by the 300 by Eco R1 fragment is a separatemessage. Alternatively, there could be just one 0.8 kb messagewhich lies proximal to the 1.2 but still encroaches into the 1.5 kbHind III/Eco RI fragment. Although this analysis does not allow usto distinguish between these two possibilities, the 1.2 kbtranscript lying immediately adjacent to the P element insert is ofparticular interest.The 1.5 kb Eco RI/ Hind III fragment, which is proximal to theP insert, was used as a probe to screen a Drosophila disc cDNAlibrary. Out of 250,000 bacteriophage screened, eight positiveplaques were chosen at random and purified by successive rounds ofamplification and screening. These cDNAs range in size from 0.8 kbto 1.2 kb (Figure 20a) and cross-hybridize with the 1.5 kb EcoRI/Hind III fragment which was used to screen the library (Figure20b) as well as the largest cDNA: cDNA 6 (Figure 20c).cDNA 6 was used as a probe to hybridize to Poly(A)+ wild-typeRNA from 0-6 hour embryos (Figure 21). Since this cDNA only151Figure 20:^Putative mfs(2)31C.^cDNA6. cDNAs hybridized with^B. pH1.5 andLane 1 X Hind III2 cDNA1 Eco RI3 cDNA2 Eco RI4 cDNA3 Eco RI5 cDNA4 Eco RI6 cDNA5 Eco RI7 cDNA6 Eco RI8 cDNA7 Eco RI9 cDNA8 Eco RIA1 2^3 4^5 6 7^8^9B1^2^3^4^5^6^7^8^9C152Figure 21: Poly(A)+ RNA from 0-6 hour Oregon-R embryos hybridizedwith cDNA 6.1.2 kb153hybridizes to a 1.2 kb transcript, it is probable that the 0.8 andthe 1.2 kb messages, detected by the 1.5 kb Eco RI/Hind IIIfragment, are adjacent transcripts rather than derivatives of thesame transcription unit.To determine if the 1.2 kb transcript which corresponds tocDNA 6 is disrupted in mfs(2)31 mutant strains, Poly(A)' RNA wasisolated from all four alleles as well as the revertant strainsmfs Rl and mfsR2. The RNA was hybridized to a probe made from thefull length cDNA 6 (Figure 22a) and then the filter was strippedand reprobed with pDmA2 to verify equal loading (Figure 22b). Inall cases but one, there is no significant alteration in levels oftranscription of the 1.2 kb message nor are any additional messagesdetected. The single exception was in the mfs2 P elementcontaining strain in which the level of transcription wassignificantly reduced. Since this reduction is not observed ineither of the two revertant strains, I conclude that the P elementin subdivision 31E is interfering with the expression of theadjacent 1.2 kb transcript.To observe the developmental profile of the 1.2 kb transcript,Poly(A)' RNA was isolated from a number of developmentally stagedDrosophila tissues: 0-2.5 hr embryos, 2.5-6 hour embryos, 6-9 hourembryos, 9-12 hour embryos, 12-24 hour embryos, 1st instar larvae,2nd instar larvae, third instar larvae, early pupae, late pupae andadult males and females. Following hybridization with a cDNA 6probe (Figure 23a), the filter was stripped and re-probed withpDmA2 (Figure 23b). The hybridization with pDmA2 reveals that154Figure 22:^Poly(A)* RNA from the mfs(2)31 alleles hybridized withA. cDNA 6^B. pDmA2.^RNA is from 0-6 hr embryos.Lane 1^Oregon-R2 mfs3 b pr on/Q/03 mfs4 b pr cn/CyO4 mfs2 b pr cn/CyO5 mfs1 b pr cn/CyO6 mfs-RI b pr on/Cy-07 mfsR2 b pr cn/C370A1^2^3^4^5^6^7411 0 SS smoBeillaa it*155Figure 23: Developmental profile of a putative mfs(2)31transcript. Staged poly(A)+ RNA hybridized with A.cDNA 6 B. pDmA2.^1. 0-3 hr 2. 3-6 hr 3. 6-9 hr4. 9-12 hr 5. 12-24 hr 6. 1st Instar 7. 2ndInstar 8. 3rd Instar 9. Early pupae 10. Latepupae 11. Adult male 12. Adult female.A1 2 3 4 5 6 7 8 9 10 11 12156there is a great deal of variability in the loading of this gel andthus, little can be said about the relative levels of expression ofthe 1.2 kb transcript. Nonetheless, it is clear that a single 1.2kb message is expressed in all stages of development including theadults of both sexes. The expression in the larvae and the adultsis consistent with the larval lethal period and adult functionwhich have been demonstrated for mfs(2)31 (see Chapter 3).Sequencing mfs(2)31 Since cDNA 6 hybridizes to sequences adjacent to the site ofthe P element insertion and is the most likely candidate for thetranscription unit which encodes mfs2, it was sequenced in itsentirety. Initially, the ends of all 8 cDNAs were sequenced todetermine their alignment. All have identical 3' ends but differat their 5' ends (data not shown) which indicates that the shortercDNAs are probably not full length. cDNA 6 is 1230 by long (Figure24) and contains a single long open reading frame which encodes aprotein 259 aa in length (30 kDa). There is a 400 by untranslated5' leader which contains numerous stop and start codons in allthree reading frames, and a short 3' untranslated sequence (70 bp)which contains both a polyadenylation signal and a poly(A) tail.The 30 Kd protein predicted by the cDNA 6 sequence is quitebasic with 66 of the 259 residues being either lysines orarginines. Beginning at the amino terminus, the basic residues aredispersed somewhat randomly throughout the first 180 aa. Theprotein sequence also contains nine potential phosphorylation sites157Figure 24: The complete sequence of cDNA 6. The amino acidsequence of the mfs(2)31 gene is presented above thenucleotide sequence beginning at +384 and ending at+1160. The polyadenylation signal is underlined andpotential sites of phosphorylation are indicated byasterisks.AAGGTAAGTTTTTGGGGTTGTTACATTTTAAGCATCGGTTAAGTGAAAATAAACATACAAATACATACTTAAAGTAGTCCTTAACACTTG^90TAAACTCTTACGAGTGAAATACATGACAATTGTTGATTGAACAATGAGAATGATTGCAGTGCACATGTAAAATTGACAAGCAATCAAAAT 180TGCTAAAAGATACCCGTGAACAAAAAAAATATTGTGACATTTAAATTTATACATTAAGGTAAGTGAATTTCCGCGTGATTAGTTATCTGT 270GGATGGCCACAGAGGCTTCCCCAGGAAACAAATTTCGCACACCCCCAACCGTCCACATGTTAATAACCACATGAATGTATTTCAGATCTG 360MPQNEYMER H R K L Y G R R L DY EEGTCTACAGAATTAACAATTAACAATGCCGCAGAATGAGTATATCGAACGCCATCGCAAGCTGTATGGCCGGCGATTGGATTACGAGGAAC 45023^R K R K K E A R L P K DR A R K A R K L R G I K A K L F N KGGAAACGCAAGAAGGAAGCGCGTCTTCCCAAGGACCGAGCACGAAAGGCTCGCAAGTTGCGCGGCATCAAGGCCAAACTCTTTAATAAGG 54053^ER R N E K IQ I K K K I Q A H E E K K V K K Q E E K V E DAGCGACGCAATGAAAAGATTCAGATTAAGAAGAAGATCCAGGCCCACGAAGAGAAGAAGGTCAAAAAGCAGGAGGAGAAGGTCGAGGATG 63083^GA L P H Y L L D R G I Q S S A K V L S N M I K Q K R K E KGAG CCCTG CCGCATTATCTG CTCGACAGAGGCATCCAC TCCAGCG CCAAGGTCCTGTCCAATATGATCAAG CAGAAG CGCAAGGAGAAGG 720113^A G K W D V P I P K V R A Q S D A E V F K V L K T G K T K RCAGGCAAGTGGGACGTGCCCATTCCCAAAGTACGCGCTCAGTCAGATGCTGAGGTCTTCAAAGTACTAAAGACCGGAAAGACAAAGCGAA 810143^K A W K R M V T K V T F VG E N F T R K P P K F ER F I R PAGGCATGGAAG CG CATGG TCACAAAAG TCACATTCGTCGGTGAGAACTTCACACGCAAGCCACCAAAGTTCGAGCGTTTCA TTCGACC CA 900173^MG L R M K K A H V T H P EL K A T F N L P I I G V K K N PTGGGTCTGCGCATGAAGAAGGCTCACGTTACGCATCCAGAACTGAAAGCCACCTTCAATCTGCCCATCATTGGCGTCAAGAAGAACCCCA 9902 0 3^SS P R F T S L G V I T K G TV I E V N I S E L G L V T Q TGCTCGCCCAGGTTCACTTCCTTGGCTGTAATTACAAAGGGTACTGTGATCGAAGTCAACATCTCTGAGCTGGGTTTGGTAACGCAAACGG 10802 33^G K V V W G K Y A QV T N N P E N D G V IN A V L L V*GAAAAGTTGTCTGGGGCAAATACGCTCAGGTCACGAACAATCCAGAAAACGATGGTGTTATCAATGCAGTGCTGCTTGTCTAACCCCGAA 1170AA TG TCGCTACCTAATTTAAGAG TTTGATTTAATAAACAAACATATACAATTTTTTTTTT^ 1230158for either protein kinase C or casein kinase II, which are randomlydispersed throughout its length. Although this protein has nosimilarities with any previously described Drosophila genes, acomparison of the predicted protein sequence with sequences in theEMBL/Genbank data base suggests that cDNA 6 may encode amicrotubule associated protein (Figure 25). There is 26%similarity between cDNA 6 and mouse microtubule associated protein1B (MAP1B) over a 130 amino acid stretch. The similarity withMAP1B is in a very basic domain which contains many copies of thesequence KKEEand KKEI/V at randomly dispersed intervals. Sincethis segment is responsible for the binding of MAP1B tomicrotubules (Noble et a/., 1989), it may indicate an equivalentfunction for the cDNA 6 encoded protein. Alternatively, since thetwo sequences are both rich in basic residues, this similarity maysimply be fortuitous.Although there is no similarity between the cDNA 6 encodedprotein and any non-histone chromosomal proteins, it has mildsimilarity (22% over 60 aa) in its carboxy terminus with themammalian histone H1 protein. Even though histone H1 is also avery basic protein, this similarity is not attributable to randommatches of basic residues. Although this similarity is somewhatlimited, it is notable because of the su(var) phenotype of mfs(2)31mutations.To determine the location of the P element insert with respectto the transcription unit identified by cDNA 6, the ends of the 1.0kb Hind III (pH1.0) and the 1.5 kb Hind III/Eco RI (pH1.5) genomic159Figure 25: Similarities with the deduced protein sequence ofmfs(2)31. In each case, the upper sequence correspondsto mfs(2)31. Perfect matches are indicated by doubledots while conserved changes are indicated by singledots.MOUSE MICROTUBULE -ASSOCIATED PROTEIN: MAP1B26.2% identity in 130 aa overlap30^40^50^60^70^80^90KRKKEARLPKD-RARKARKLRGIKAKLFNKERRNEKIQIKKKIQAHEEKKVKKQEEKVEDGALPH• • •^ • •^• •^• •^•^• • •^•KKEDKTPLKKDEKPRKEEVKKEIKKEIKKEERKELKKEVKKETPLKDAKKEVKKEEKKEVKK---670^680^690^700^710^720100^110^120^130^140^150YLLDRGIQSSAKVLSNMIKQKRKEKAGKWDVPI-PKVRAQSDAEVFKVLKTGKTKRKAWKRMVTK• • • • •^ • •^•^•--EEKEPKKEIKKISKDIKKSTPQSDTKKPSALKPKVAKKEESTKKEPLAAGKLKDKGKVKVIKK730^740^750^760^770^780MAMMALIAN HISTONE21.7% identity in 60 aa overlap170^180^190^200^210PPKFERFIRPMGLRMKKAHVTHPELKATFNLPIIGVKKNPSSPRFTSLGVITKGTVIEVN:^........... . : .PPVSELITKAVAASKERSGVSLAALKKALAAAGYDVEKNNSRIKLGLKSLVSKGTLVQTK40^50^60^70^80160fragments were sequenced. The Hind III site, which is shared bythese adjacent fragments, corresponds to the internal Hind III sitewhich is found 75 by from the end of intact P elements. Thus, theHind III end of the 1.5 kb fragment begins with the terminal 75 byof a P element sequence. Beginning 37 by downstream of thiselement are sequences which are identical to the 5' end of cDNA 6(Figure 26). Partial sequencing of the 1.0 kb Hind III fragmentrevealed that the P element is approximately 600 by in length andcontains all of the sequences which are considered necessary fortransposition (Figure 27). Although the element has not beensequenced in its entirety, it contains an internal tract of A's andT's, at least 85 by in length, which is not P element sequence.To determine the structure of the mfs(2)31 transcription unit,a wild-type copy of the gene is required. The 1.5 kb Hind III/EcoRI fragment from pH1.5 was used as a probe to screen a cosmidlibrary. Out of 25,000 colonies screened, a single cosmid wasisolated and purified: cos50. If cos50 DNA is digested with Bam HIand Sal I and hybridized with the 1.5 kb Hind III/Eco RI fragmentused to screen the library, the pattern of hybridization is thesame as what would be predicted from a digestion of X1 (compareFigure 14 with Figure 28). In addition, the hybridization of aseries of overlapping fragments which span X1 with a probe madefrom the complete cos50 construct indicates that these two clonesare collinear (Figure 29). A 1.9 kb Eco RI/Hind III fragment,which hybridizes to cDNA 6, was subcloned out of cos50 into pUC19.This fragment, which was sequenced in its entirety (Figure 30),161Figure 26: The relationship between the P element insert and the 5'end of cDNA 6 in mfs2 . Compressed lettering indicatessequences which are identical to the 5' end of cDNA 6.P element sequences are underlined while doubleunderlining indicates a reasonably good (5/7) insect capsite.AGCTTCGGCT TTCGACGGGA CCACCTTATG TTATTTCATC ATGGTGTGTT CTTTCTTTCCTTTGATTCTG TTGCCGAAGC AAGGTAAGTTTTTGCGGTTGTTACATTTTAAGCATCGGTT162Figure 27: Partial sequence of the 1.0 kb Hin dIII from theplasmid pH1.0. Single underlining indicates P elementsequences while double underlining indicates the non Pelement sequences which are within the insert.CAGGCATCAAAACAAATTATGCATCGCTTCTACGGTTTTTATTTTATAATTTATGTTCCAAACTCCGAATCATGATGAAACAATAAGTGCGCTTGTCAGGCCTAGAACTTTACTTTTGTGAAATTTTTATTTTTGTAATATTCCACGCGTGCAGCCCTGCTAACATAAGGGAGTGAAAGGAGTGCTACCTAAATTTTTATCTTGGAAAAATTTTTGTTTCAGGAAAAGTGTCCAAAAATAGCTCAATTTTTGGTCCCGTCAATAGTATTCGGATTTGTTAACATGTTTTTGTATATGCACCTTTATGATGTATTGGATATAACGCGTAGTTCAACTTCCTGGCAAGAGACTGAGTGTCGTTTTTAATTTGATTTAATAGGATTTTGGACCATTTTTTTTTGGCTTCTGGTTCAAAATATAGTCACGCTGTATCCACTTAAATTGAGTCTGTAAGAGTATTCGGATTTGGAACGAAAATTTGATTTGTTATTCATCCCAGTAACAAAAGTAGTGTGTGTTCCGTATGCTTG AGTGAGACAGC^^ATATAA AATTATAAATT TTTATAAAAT TATAAAATTA TATATTATAA TATATTATAA AATATATTTAT ATTATATTA AAATATATTTATATTTCATT TTTTTTTATT CCACGTAAGG GTTAATGTTTT CAAAAAAAA ATTCGTCCGC ACACAACCTT CCTCTCAACA AGCAAACGTG CACTGAATTTA AGTGTATAC TTCGG 163Figure 28: cos50 DNA hybridized with pH1.5.Lane^1^X^Hind III2^cos50^Bam HI3^cos50 Sal IA. Photograph of ethidium bromide stained gel.B. Photograph of autoradiogram.A^ B1 2 3^ 1^2^3 9.5 kb8.5 kb164Figure 29: Overlapping fragments from X1 hybridized with cos50.Lane^1 X^Hind III2 pB2.8^Bam HI3 pS5.0 Sal I4 pP6.0^Pst I5 pR8.5 Eco RI6 pB2.8^Bam H1A. Photograph of ethidium bromide stained gel.B. Photograph of autoradiogram.pUCMID 165Figure 30: The complete genomic sequence of the region encompassingthe gene represented by cDNA 6. The sequencecorresponding to cDNA 6 is underlined.TAGAG TCGACCTG CAGGCATCAAG CTTG TCAGGAGTGCTACCTGGATTTG TTATTTTAATTTGTAAGAGTATTAACAAATTATCCTAGAA^90CTTAAATTTGTATACATGTTTTTATTTAATAGGCGGATTTGGAGCATCGCTTCTACTTTTGTGCTTGGAAAAAGTATATGCACATTTTGG^180ACCACGAAAATTTTACGGTTTTTAAATTTTTATTTTTTGTTTCCTTTATGATGATTTTTTTTTGATTTGTTATATTTTATAATTTTTGTA^270ATAAGGAAAAG TGTATTGGATATGG CTTCTGGTTCATCCCAG TTTATGTTCCATTCCACGCGTTCCAAAAATAAACG CGTAG TTCAAAAT^360ATAAACAAAAGTAAACTCCGAATGCAGCCCTGCGCTCAATTTTTCAACTTCCTGTCACGCTGTGTGTGTGTTCTTTCTTTCCTTTGATTC^450TGTTGCCGAAGCAAGGTAAGTTTTTGGGGTTGTTACATTTTAAGCATCGGTTAAGTGAAAATAAACATACAAATACATACTTAAAGTAGT^540CCTTAACACTTGTAAACTCTTACGAGTGAATACATGACAATTGTTGATTGAACAATGAGAATGATTGCAGTGCACATCTAAAATTGACAA^630GCAATCAAAATTGCTAAAAGATACCCGTGAACAAAAAAAATATTGTGACATTTAAATTTATACATTAAGGTAAGTGAATTTCCGCGTGAT^720TAGTTATCTG TGGATGGCCACAGAGGCTTCCCCAGGAAACAAATTTCGCACAACCGTCCACATGTTAATAACCACATGAATG TATTTCAG^810ATCTGGTCTACAGAATTAACAATTAACAATGCCGCAGAATGAGTATATGGAACGCCATCGCAAGCTGTATGGCCGGCGATTGGATTACGA^900GGAACGGAAACGCAAGAAGGAAGCTCGTCTTCCCAAGGACCGAGCACGAAAGGCTCGCAAGTTGCGCGGCATCAAGGCCAAACTCTTTAA^990TAAGGAGCGACGCAATGAAAAGATTCAGATTAAGAAGAAGATCCAGGCCCACGAAGAGAAGAAGGTCAAAAAGCAGGAGGAGAAGGTCGA 1080GGATGGAGCCCTGCCGCATTATCTGCTCGACAGAGGCATCCAGTCCAGCGCCAAGGTCCTGTCCAATATGATCAAGCAGAAGCGCAAGGA 1170GAAGGCAGGCAAGTGGGACGTGCCCATTCCCAAAGTACGCGCTCAGTCAGATGCTGAGGTCTTCAAAGTACTAAAGACCGGAAAGACAAA 1260GCGAAAGCCATGGAAGCGCATGGTCACAAAAGTCACATTCGTCGGTGAGAACTTCACACGCAAGCCACCAAAGTTCGACCGTTTCATTCG 1350ACCCCTGGGTCTGCGCATGAAGAAGGCTCACGTTACGCATCCAGAACTTAAAGCCACCTTCAATCTGCCCATCATTGGCGTCAAGAAGAA 1440CCCCAGCTCGCCCATGTTCACTTCCTTGGGTGTAATTACAAAGGGTACTGTGATCGAAGTCAACATCTCTGAGCTGGGTTTGGTAACGCA 1530AACGGGAAAAGTTGTCTGGGGCAAATACGCTCAGGTCACGAACAATCCAGAAAACGATGGTGTTATCAATGCAGTGCTGCTTGTCTAACC 1620CCGAAAATGTTGCTACCTAATTTAAGAGTTTGATTTAATAAACAAACGTATACAATAACATGAACCATCGTTTGATTCTCTTAGCTTATT  1710ATACTATTAGCTGTAGCAATGCCATATATATAAACACACATATCTATACCAACATCTTTTTGTTTCTTGCAGTTTGGTGACTGGAATACA 1800AGATGCAGATCTTCGTTAAGCATCACGGGCAAGACCATCACTCTTGAGGTCGAACCGTCGGATACTATCGAAAATGTCAAAGCCAAGATT 1890CAGGACAAAGAAGGAATTCGAGCTCGGTACCCGGGGATCCTCTAGAGTCGAC^ 1942166corresponds to the 1.5 kb Eco RI/Hind III and 1.0 kb Hind IIIfragments from X1 (minus the P element insert). A comparison ofthe sequences of these three fragments indicates that the P elementinsertion is not associated with any rearrangement of genomic DNA.Comparing the sequence of cDNA 6 with this 1.9 kb fragmentindicates that the transcription unit does not contain any introns.The 5' end of the gene was examined for potential regulatorysequences (Figure 31). Although there is a good consensus sequencefor an insect cap site at -16 from the 5' end of cDNA 6 (ATTCTGTT),the closest TATA box is at -103 (TATAAA). The TATAAA sequencefalls within an imperfect (13/15) 15 by direct repeat whosefunction is not known. If these are the promoter sequences ofmfs(2)31, then cDNA 6 may not be full length and the P element mayhave inserted into the 5' untranslated leader sequence of thetranscription unit. Although mfs2 is a hypomorphic allele and thusshould be producing some functional message, a larger transcripthas not been detected in the transcriptional analysis. It isformally possible that there is a small intron, 5' to cDNA 6, intowhich the element has inserted. There are several splice donorsites immediately 5' to the element and two splice acceptor sitesbetween the site of the insert and the 5' end of cDNA 6. If thistranscription unit contains an additional 5' exon it would have tobe small since the size of the message is virtually identical tothe size of cDNA 6. Since the 5' untranslated leader containsnumerous stop and start codons in all three frames, it is unlikelythat such an exon would be coding. It is more plausible that cDNA167Figure 31: Potential upstream regulatory sequences of mfs(2)31. Thesequence corresponding to cDNA 6 is indicated bycompressed lettering. Single underlining indicates animperfect (13/15) 15 by direct repeat which contains theclosest TATA box (TATAA). Double underlining indicatesa reasonably good (5/7) insect cap site consensussequence. An asterisk indicates the site of the Pelement insertion.GCAGGCATCA AGCTTGTCAG GAGTGCTACC TGGATTTGTTTAACAAATTA TCCTAGAACT TAAATTTGTA TACATGTTTTAGCATCGCTT CTACTTTTGT GCTTGGAAAA AGTATATGCATTACGGTTTT TAAATTTTTA TTTTTTGTTT CCTTTATGATTATTTTATAA TTTTTGTAAT AAGGAAAAGT GTATTGGATATTTATGTTCC ATTCCACGCG TTCCAAAAAT AAACGCGTAGATTTTAATTT GTAAGAGTATTATTTAATAG GCGGATTTGGCATTTTGGAC CACGAAAATTGATTTTTTTT TGATTTGTTATGGCTTCTGG TTCATCCCAGTTCAAAATAT AAACAAAAGTAAACTCCGAA TGCAGCCCTG CGCTCAATTT TTCAACTTCC TGTCACGCTG TGTGTGTGTTCTTTCTTTCCTTTGATTCTGTTGCCGAAGCAAGCTAAGTTTTTGGGGTTGTTACATTTTAAGCATOGGTTAAGTGAAAATAAAC1686 is full length and that the P element has inserted 5' to the siteof transcriptional initiation. In doing so, it may be interferingwith the regulatory sequences of this transcription unit.169DiscussionThis chapter describes the use of a P element induced mutationto clone and sequence the mfs(2)31 locus of Drosophilamelanogaster. There are several reasons why we are confident thatwe have identified the correct transcription unit. First, we haveshown that the cloned P element is responsible for the mfs(2)31phenotype associated with the dysgenic allele mfs 2. Hybriddysgenesis induced revertants of this allele are invariablyassociated with the excision of the cloned element. Second, theelement has inserted immediately 5' to the transcriptional startsite of a 1.2 kb message and reduces the level of this transcript.A transcriptional analysis of the region has failed to detect amessage within 7 kb of the proximal end of the insert. Finally,although it is formally possible that the P element is interferingwith a second, more distant message, the cytogenetic analysisdescribed in chapter 2 provides no evidence that the mfs2 allelemay be a double mutation. Although the 1.2 kb message is a strongcandidate for the mfs(2)31 transcription unit, a formal proof wouldrequire either the sequencing of other mfs(2)31 alleles or a rescueof the mutant phenotype by P element mediated germlinetransformations. Additional mfs(2)31 mutant alleles are beingsequenced to verify the identity of this transcription unit.A comparison of cDNA sequences with the corresponding genomicsequences from a wild-type and the mfs 2 strain indicate that the Pelement has inserted 37 by upstream of the transcriptional startsite. Although this localization is based on the assumption that170the longest cDNA isolated (cDNA 6) is full length, there areseveral reasons why we feel that this assumption is correct.First, cDNA 6 is identical in size to the only transcript detectedat this locus. Second, a transcriptional analysis in the regionhas failed to detect any messages within seven kb of the distalside of the insert. Finally, the apparent insertion of the elementinto 5' regulatory sequences is consistent with the transcriptionalprofile of mfs 2 b pr cn/CyO heterozygotes. Although the amount ofthe 1.2 kb message is substantially reduced in this strain, noaberrant messages have been detected. This is particularlysignificant since the junctions of the mfs 2 insertion do notconform to the eukaryotic splice consensus sequence making itunlikely that the element would be precisely excised during RNAprocessing.The insertion of a P element into the regulatory region ofmfs 2 is consistent with the genetic analysis of this mutation whichshows that it has reduced, but not a zero level of function.However, one would not necessarily predict that a regulatorymutation would have a temperature-sensitive phenotype. Anexamination of the sterility associated with mfs 1 tmfs 2transheterozygotes revealed that the phenotype is temperaturesensitive with a sensitive period which extends into adulthood.Based on the transcriptional analysis of these respective alleles,it now seems more likely that this temperature-sensitivity isattributable to the mfs l mutation. The pattern of expression ofthe 1.2 kb transcript is unaltered in the mfs l mutation, which171makes it more likely that it contains a lesion in the coding regionof the gene.The mfs(2)31 transcript is present at all stages ofdevelopment as well as in adults of both sexes. The presence ofthe mfs(2)31 transcript in the larvae, pupae and adults wasanticipated from the previous genetic characterization of the locussince zygotic mfs(2)31 functions have been demonstrated during allof these stages of the Drosophila life cycle (Chapter 3). Incontrast, the presence of an embryonic mfs(2)31 transcript issomewhat surprising since the larval stage is the earliest periodfor which we have demonstrated an essential function for the locus.It is possible that the gene product does not have an essentialfunction during embryonic development or alternatively, theembryonic function may be under maternal control. This latterpossibility is supported by our detection of large amounts ofmfs(2)31 transcripts in 0-2.5 hr embryos, transcripts which areunlikely to have been the product of zygotic transcription(Anderson and Lengyel, 1979; Edgar and Schubiger, 1986).A comparison of mfs(2)31 genomic and cDNA sequences indicatethat the structure of this locus may be fairly simple. There is asingle message detected at this locus and there are no introns ineither the protein coding sequences or the 5' untranslated leader.Although the lack of complexity of the mfs(2)31 locus at themolecular level appears to be inconsistent with the manydevelopmental functions of the gene, it should be noted that theanalysis of the four mutant alleles has suggested little, if any,172genetic complexity at this locus (Chapter 3). These mutations fitinto a relatively straightforward allelic series of increasedseverity. Hypomorphic alleles which exhibit bristle and sterilityphenotypes, also exhibit the essential larval function ashemizygotes which is observed in the homozygotes of strongeralleles. Thus, while the mfs(2)31 locus clearly is required for avariety of developmental events, our molecular analysis suggeststhat these various functions may be mediated by the same geneproduct.What are these various functions of the mfs(2)31 locus? Thebest described of all its functions is its requirement for thematuration of the adult reproductive systems. Male sterility hasbeen correlated with a number of spermatid defects, all of whichcan be explained by abnormal centriole behavior (Lindsley et a/.,1980), while female sterility has been correlated with a delay inthe onset and completion of vitellogenesis (Chapter 3). mfs(2)31also has a pupal function necessary for bristle elaboration and anessential function during larval development. We have beenprimarily interested in mfs(2)31 because of its su(var) phenotype.Surviving allelic combinations suppress position-effect variegationin a variety of variegating backgrounds. It has long been thoughtthat genes which can be mutated to modify PEV will encode proteinswhich are required for the assembly and/or maintenance of chromatinstructure.We have previously proposed that the variety of phenotypesassociated with mfs(2)31 mutations are consistent with a defect in173chromosome behavior, either during or subsequent to cellulardivision (see Chapter 3). Such a defect could not only interferewith heterochromatin assembly, thus producing a su(var) phenotype,but it could also prevent a cell from achieving the maximum ratesof protein synthesis which are required for bristle elaboration,vitellogenesis and embryonic development, since all of theseprocesses depend on an expedient switch from a mitotic to anendomitotic cycle. A defect associated with nuclear architecturewould also account for the abnormal nuclear morphology andcentriole behavior observed during spermatid development.The predicted amino acid sequence for the cDNA 6 encodedprotein is not particularly informative with regard to specificbiochemical activities that the mfs(2)31 product may have.However, it is intriguing that the only meaningful similarity wehave discovered for this protein is to the predicted product of themouse MAP1B gene (Noble et al., 1989). MAP1B encodes a microtubuleassociated protein which has been identified in mouse brain tissue.Although the function of brain MAPs is essentially unknown,phosphorylated MAP1B may play a role in the cytoskeletal changesthat accompany neurite extension. MAP1B is thought to bind to atleast two tubulin subunits in the microtubule polymer, and thisbridging of subunits might be involved in either nucleatingmicrotubule polymerization or the stabilization of the structure.The region which is responsible for the binding of MAP1B tomicrotubules is a highly basic sequence with many copies of themotifs KKEE and KKEI/V which are repeated but not at fixed174intervals. This is the region with which the highly basic mfs(2)31protein product shares its greatest similarity. Although thesimilarity between MAP1B and mfs(2)31 may be fortuitous, it istempting to consider that the sequence similarity could be relatedto the fact that both genes are involved in binding tomicrotubules, a function which would be consistent with a role forthe mfs(2)31 product in the control of chromosomal movement andstructure. A possible nuclear function for the mfs(2)31 geneproduct related to chromosomal behavior is further suggested by aweak homology in the carboxy terminal of the protein with mammalianhistone H1, a protein which is required for the condensation of the10 nm chromatin fiber into the more compact 30 nm structure.It has been proposed that dominant Su(var) mutations willidentify genes which encode structural components ofheterochromatin. It would seem apparent that any protein whichfunctions in some aspect of chromosomal behavior could be requiredfor expedient chromatin assembly, and thus could be mutated tomodify PEV. Such genes would differ from the more typicalmodifiers of PEV in that they are unlikely to exhibit dominantphenotypes. We feel that it is possible that the mfs(2)31 locusencodes a protein which associates with and/or controls thebehavior of chromosomes during the cell cycle. If so, it mayextend the usefulness of PEV from an assay for geneticallydissecting chromatin, to an assay for genetically dissecting thenuclear architecture.175LITERATURE CITEDAlfageme, C.R., G.T. Rudkin, and L.H. Cohen (1980). Isolation,properties and cellular distribution of D1, a chromosomal proteinof Drosophila. Chromosoma 78, 1-31.Ananiev, E.V., and V.A. Gvozdev (1974). Changed pattern oftranscription and replication in polytene chromosomes of Drosophilamelanogaster resulting from eu-heterochromatin rearrangement.Chromosoma 45, 173-191.Anderson, K.V. and J.A. Lengyel (1979). Rates of synthesis of majorclasses of RNA in Drosophila embryos. Dev.Biol. 70, 217-231.Annunziato, A.T., L.Y. Frado, R.L. Seale, and C.L.F. Woodcock(1988). Treatment with sodium butyrate inhibits the completecondensation of interphase chromatin. Chromosoma 96, 132-138.Ashburner, M. (1989). Drosophila-A Laboratory Handbook. Cold SpringHarbor Laboratory, Cold Spring Harbor, New York.Ashley, C.T., C.G. Pendleton, W.W. Jennings, A. Saxena, and C.V.C.Glover (1989). Isolation and sequencing of cDNA clones encodingDrosophila chromosomal protein Dl. J. Biol. Chem. 264, 8394-8401.Baker, W.K. (1967). A clonal system of differential gene activityin Drosophila. Dev. Biol. 16, 1-17.Baker, W.K. (1968). Position-effect variegation. Adv. Genet. 14,133-169.Barigozzi, C., S. Dolfini, M. Fraccaro, C.R. Raimondi, and L.Tiepolo (1966). In vitro study of the DNA replication patterns ofsomatic chromosomes of Drosophila melanogaster. Exp. Cell Research43, 231-234.Beckingham, K., and A. Rubacha (1984). Different chromatin statesof the intron - and type 1 intron' rRNA genes of Calliphoraerythrocephala. Chromosoma 90, 311-316.Bell, A.E. (1954). A gene in Drosophila melanogaster that producesall male progeny. Genetics 39, 958-959.Belote, J.M., F.M. Hoffman, M. McKeown, R.L. Chorsky, and B.S.Baker (1990). Cytogenetic analysis of chromosome region 73AD ofDrosophila melanogaster. Genetics 125, 783-793.Benton, W.D., and R.W. Davis (1977). Screening 2gt recombinantclones by hybridization to single plaques in situ. Science 196,180-182.176Biessmann, H., P. Kruger, C. Schroper, and E. Spindler (1981).Molecular cloning and preliminary characterization of a Drosophilamelanogaster gene from a region adjacent to the centromeric p-heterochromatin. Chromosoma 82, 493-503.Boffa, L.C., R.J. Gruss, and V.A. Allfrey (1981). Manifold effectsof sodium butyrate on nuclear functions. J. Biol. Chem. 256, 9612-9621.Brennan, M.D., A.J. Weiner, T.J. Goralski, and A.P. Mahowald(1982). The follicle cells are a major site of vitellogeninsynthesis in Drosophila melanogaster. Develop. Biol. 89, 225-236.Brock, J.K (1989). A genetic analysis of region 31 on chromosome 2of Drosophila melanogaster. M.Sc. Thesis, University of BritishColumbia.Brosseau, G.E. Jr. (1970). V-type position effects for e* and ro+in Drosophila. Drosophila Inform. Serv. 45, 100.Candido, E.P.M., R. Reeves, and J.R. Davie (1978). Sodium butyrateinhibits histone deacetylation in cultured cells. Cell 14, 105-113.Catcheside, D.G. (1938). A position-effect in Oenothera. J. Genet.38, 345-352.Catcheside, D.G. (1947). The P-locus position effect in Oenothera.J. Genet. 48, 31-42.Cattanach, B.M. (1974). Position-effect variegation in the mouse.Genet. Res. Camb. 23, 291-306.Caudy, M., H. Vassin, M. Brand, R. Tuma, L.Y. Jan, and Y.N. Jan(1988). daughterless, a Drosophila gene essential for bothneurogenesis and sex-determination, has sequence similarities tomyc and the achaete-schute complex. Cell 55, 1061-1067.Christman, J.K., N. Weich, B. Schoenbrun, N.K. Schneiderman, and A.Acs (1980). Hypomethylation of DNA during differentiation of Frienderythroleukemia cells. J. Cell Biol. 86, 366-370.Clegg, N.J. (1992). Suppressors of position-effect variegation andthe cdc2Dm gene in Drosophila melanogaster. PhD. Thesis, Universityof British Columbia.Cline, T.W. (1989). The affairs of daughterless and the promiscuityof developmental regulators. Cell 59, 231-234.Clutterbuck,^A.J.,^and D.H.^Spathas^(1984).^Genetic andenvironmental modification of gene expression in the br1Al2variegated position effect mutant of Aspargillus nidulans. Genet.Res. 43, 123-138.177Cohen, J. (1962). Position effect variegation at several closely-linked loci in Drosophila melanogaster. Genetics 47, 647-659.Cronmiller, C., P. Schedl, and T.W. Cline (1988). Molecularcharacterization of daughterless, a Drosophila sex determinationgene with multiple roles in development. Genes Dev. 2, 1666-1676.De Valoir, T., M.A. Tucker, E.J. Belikoff, L.A. Camp, C. Bolduc,and K. Beckingham (1991). A second maternally expressed Drosophilagene encodes a putative RNA helicase of the "DEAD box" family.Proc. Natl. Acad. Sci. USA 88, 2113-2117.Demerec, M., and H. Slizynska (1937). Mottled white 258-18 ofDrosophila melanogaster. Chromosoma 33, 319-344.Devlin, R.H., B. Bingham, and B.T. Wakimoto^(1990). Theorganization and expression of the light gene, a heterochromaticgene of Drosophila melanogaster. Genetics 125, 129-140.Dimitri, P., and C. Pisano (1989). Position-effect variegation inDrosophila melanogaster. Relationship between suppression effectand the amount of Y chromosome. Genetics 122, 793-800.Dorbic, T., and B. Wittig (1987). Chromatin from transcribed genescontains HMG17 only downstream from the starting point oftranscription. EMBO J. 6, 2393-2399.Dorn, R., S. Heymann, R. Lindigkeit, and G. Reuter (1986).Suppressor mutation of position-effect variegation in Drosophilamelanogaster affecting chromatin properties. Chromosoma 93, 398-403Edgar, B.A., and G. Schubiger (1986). Parameters controllingtranscriptional activation during early Drosophila development.Cell 44, 871-877.Eissenberg, J.C. (1989). Position-effect variegation in Drosophila,towards a genetics of chromatin assembly. Bioessays 11, 14-17.Eissenberg, J.C., T.C. James, D.M. Foster-Hartnett, T. Hartnett, V.Ngan, and S. Elgin (1990). Mutation in a heterochromatin-specificchromosomal protein is associated with suppression of position-effect variegation in Drosophila melanogaster. Proc. Natl. Acad.Sci. USA 87, 9923-9927.Felsenfeld, G., and J.D. McGhee (1986). Structure of the 30 nmchromatin fiber. Cell 44, 375-377.Feinberg, A.P., and B. Vogelstein (1983). A technique forradiolabelling DNA restriction endonuclease fragments to highspecific activity. Anal. Biochem. 137, 266-267.178Foe, V.E., and B.M. Alberts (1985). Reversible chromosomecondensation induced in Drosophila embryos by anoxia: visualizationof the interphase nuclear organization. J. Cell Biol. 100, 1623-1636.Frasch, M. (1991). The maternally expressed Drosophila geneencoding the chromatin-binding protein BJ1 is a homolog of thevertebrate gene Regulator of Chromatin Condensation, RCC1. EMBO J.10, 1225-1236.Fryberg, E.A., K.L. Kindle, N. Davidson and A. Sodja (1980). TheActin genes of Drosophila: a dispersed multi-gene family. Cell 19,365-378.Gall, J.G., E.H. Cohen, and M.L. Polan (1971). Repetitive DNAsequences in Drosophila. Chromosoma 33, 319-344.Giorgi, F., and P. Deri (1976). Cell death in ovarian chambers ofDrosophila melanogster. J. Embryol. Exp. Morph. 35, 521-533.Grigliatti, T. (1991). Position-effect variegation - An assay fornonhistone chromosomal proteins and chromatin assembly andmodifying factors. pp. 587-627. In, Methods in Cell Biology, Vol.35, edited by S.C.R. Elgin and B. Hamkalo, Academic Press, SanDiego.Grunstein, M. (1990). Histone function in transcription. Annu. Rev.Cell Biol. 6, 643-678.Grunstein, M., and D. Hogness (1975). Colony hybridization: Amethod for the isolation of cloned DNA's that contain a specificgene. Proc. Natl. Acad. Sci. USA 72, 3961-3965.Gowen, J.W., and E.H. Gay (1934). Chromosome constitution andbehaviour in ever-sporting and mottling in Drosophila melanogaster.Genetics 19, 189-208.Gowen, J.W., and E.H. Gay (1935). Effect of temperature on sportingeye color in Drosophila melanogaster. Science 77, 312.Gyurkovics, H., J. Gausz, J. Kummer, and F. Karch (1990). A newhomeotic mutation in the Drosophila bithorax complex removes aboundary separating two domains of regulation. EMBO J. 9, 2579-2586.Harrington M. (1990). Suppression of position-effect variegation inDrosophila melanogaster by antimorphic mutations of heterochromatinprotein components. B.Sc. Honours Thesis, University of BritishColumbia.Hartmann-Goldstein, I.J. (1967). On the relationship betweenheterochromatization and variegation in Drosophila, with special179reference to temperature-sensitive periods. Genet. Res. 10, 143-159.Hayashi, S., A. Ruddell, D. Sinclair, and T. Grigliatti (1990).Chromosomal structure is altered by mutations that suppress orenhance position-effect variegation. Chromosoma 99, 391-400.Healy, M.J., R.J. Russell, and G.L.G. Miklos (1988). Molecularstudies on interspersed repetitive and unique sequences in theregion of the complementation group uncoordinated on the Xchromosome of Drosophila melanogaster. Mol. Gen. Genet. 213, 63-71.Hearn, M.G., A. Hedrick, T.A. Grigliatti, and B.T. Wakimoto (1990).The effect of modifiers of position-effect variegation on thevariegation of heterochromatic genes of Drosophila melanogaster.Genetics 128, 785-797.Hebbes, T.R., A.W. Thorne, and C. Crane-Robinson (1988). A directlink between core histone acetylation and transcriptionally activechromatin. EMBO J. 7, 1395-1403.Heitz, E. (1934). Uber alpha and beta-heterochromatin soweikonstanz and bau der chromeren bei Drosophila. Biol. Zentralbl.54, 588-609.Henikoff, S. (1979). Position-effects and variegation enhancers inan autosomal region of Drosophila melanogaster. Genetics 93, 106-115.Henikoff, S. (1981). Position-effect variegation and chromosomestructure of a heat shock puff in Drosophila. Chromosoma 83, 381-393.Henikoff, S. (1990). Position-effect variegation after 60 years.Trends Genet. 6, 422-426.Henikoff, S., and T.D. Dreesen (1989). Trans-inactivation of theDrosophila brown gene: evidence for transcriptional repression andsomatic pairing dependence. Proc. Natl. Acad. Sci. USA 86, 6704-6708.Hess, 0. (1970). Genetic function correlated with unfolding oflampbrush loops by the Y chromosome spermatocytes of Drosophilahydei. Mol. Gen. Genet. 106, 328-346.Hessler, A.Y. (1958). V-type position effects at the light locus inDrosophila melanogaster. Genetics 43, 395-403.Hill, C.S., and J.O. Thomas (1990). Core histone-DNA interactionsin sea urchin sperm chromatin. The N-terminal tail of H2B interactswith linker DNA. Eur. J. Biochem. 187, 145-153.180Hilliker, A.J. (1976). Genetic analysis of the centromericheterochromatin of chromosome 2 of Drosophila melanogaster:deficiency mapping of EMS-induced lethal complementation groups.Genetics 83, 765-782.Hilliker, A.J., R. Appels, and A. Schalet (1980). The geneticanalysis of Drosophila melanogaster heterochromatin. Cell 21, 607-619.Hilliker, A.J., and D.G. Holm (1975). Genetic analysis of theproximal region of chromosome 2 of Drosophila melanogaster. I.Detachment products of compound autosomes. Genetics 81, 705-721.Hinton, T. (1949). The modification of the expression of a positioneffect. Am. Nat. 83, 69-94.Hinton, T., and W. Goldsmith (1950). An analysis of phenotypicreversions at the brown locus in Drosophila. J. Exp. Zool. 114,103-114.Ip, Y.T., V. Jackson, J. Meier, and R. Chalkley (1988). Theseparation of transcriptionally engaged genes. J. Biol. Chem. 263,14044-14052.Jackson, D.A. (1991). Structure-function relationships ineukaryotic nuclei. Bioessays 13, 1-10.James, T.C., and S.C.R. Elgin (1986). Identification of a non-histone chromosomal protein associated with heterochromatin inDrosophila melanogaster and its gene. Mol. Cell. Biol. 6, 3862-3872.James, T.C., J.C. Eissenberg, C. Craig, V. Dietrich, A. Hobson, andS.C.R. Elgin (1989). Distribution patterns of HP1, aheterochromatin-associated nonhistone chromosomal protein ofDrosophila. Eur. J. Cell Biol. 50, 170-180.Janning, W. (1970). Bestimmung des heterochromatisierungsstadiumsbeim white-positioneffekt mittels rontgeninduzierter mitotischerrekombination in der augenanlage von Drosophila melanogaster. Mol.Gen. Genet. 107, 128-149.Johns, E.W. (ed) (1983). The HMG Chromosomal Proteins. AcademicPress, Orlando, Florida.Johnson, L.M., P.S. Kayne, E.S. Kahn, and M. Grunstein (1990).Genetic evidence for an interaction between SIRS and histone H4 inthe repression of the silent mating loci in Saccharomycescerevisiae. Proc. Natl. Acad. Sci. USA 87, 6286-6290.Jowett, T. (1986). Preparation of nucleic acids. pp. 275-286. In,Drosophila, a Practical Approach, edited by D.B. Roberts, IRL181Press, Washington D.C..Judd, B.H. (1955). Direct proof of a variegated-type position-effect at the white locus in Drosophila melanogaster. Genetics 40,739-744.Kambysellis, M.P., and W.B. Heed (1974). Juvenile hormone inducesovarian development in cave-dwelling Drosophila species. J. InsectPhysiol. 20, 1776-1786.Karpen, G.H., and A.C. Spradling (1990). Reduced DNA polytenizationof a mini-chromosome region undergoing position-effect variegationin Drosophila. Cell 63, 97-107.Kay, M.A., and M. Jacobs-Lorena (1987). Developmental genetics ofribosome synthesis in Drosophila. Trends Genet. 3, 347-351.Khesin, R.B.,and B.A. Bashkirov (1979). Influence of deficiency ofthe histone gene-containing 38B-40 region on X-chromosome templateactivity and the white gene position-effect variegation inDrosophila melanogaster. Mol. Gen. Genet. 162, 323-328.Kimble, M., R.W. Dettman, and E.C. Raff (1990). The 03-Tubulin geneof Drosophila melanogaster is essential for viability andfertility. Genetics 126, 991-1005.King, R.C. (1970). Ovarian development in Drosophila melanogaster.Academic Press, New York.Kornher, J.S., and S.A. Kauffman (1986). Variegated expression ofthe Sgs-4 locus in Drosophila melanogaster. Chromosoma 94, 205-216.Lefevre, G. Jr. (1974). The relationship between genes and polytenechromosome bands. Ann. Rev. Genet. 8, 51-62.Lefevre, G. Jr. (1976). A photographic representation andinterpretation of the polytene chromosome of Drosophilamelanogaster salivary glands. pp. 36-61. In, The Genetics andBiology of Drosophila, Vol. la, Edited by M. Ashburner and E.Novitski, Academic Press, New York.Lefevre, G., and W. Watkins (1986). The question of the total genenumber in Drosophila melanogaster. Genetics 113, 869-895.Lehner, C.F., and P.H. O'Farrell (1990). Drosophila cdc 2 homologs:a functional homolog is coexpressed with a cognate variant. EMBO J.9, 3573-3581.Leung, J. (1988). Concerted evolution of a cluster of X-linkedtRNA4. Ser genes from Drosophila melanogaster. PhD. thesis,University of British Columbia.182Levinger, L., and A. Varshaysky (1982). Protein D1 preferentiallybinds A+T-rich DNA in vitro and is a component of Drosophilamelanogaster nucleosomes containing A+T-rich satellite DNA. Proc.Natl. Acad. Sci. USA 79, 7152-7156.Lewis, E.B., and F. Bacher (1968). Method of feeding ethylmethanesulfonate (EMS) to Drosophila males. Drosophila Inform. Ser.43, 193.Lindsley, D.L., L.S.B. Goldstein, and L. Sandler (1980). Malesterility in maternal-effect mutants. Drosophila Inform. Ser. 55,84-85.Lindsley, D.L., and E.H. Grell (1968). Genetic variations ofDrosophila melanogaster. Carnegie Inst. Wash. Publ. 627.Lindsley, D.L., and G.G. Zimm (1992). The Genome of Drosophilamelanogaster. Academic Press, San Diego.Locke, J., M.A. Kotarski, and K.D. Tartoff (1988). Dosage-dependentmodifiers of position-effect variegation in Drosophila and a massaction model that explains their effect. Genetics 120, 181-198.Mahowald, A.P., J.H. Caulton, M.K. Edwards and A. Floyd (1979).Loss of centrioles and polyploidization in follicle cells ofDrosophila melanogaster. Exp. Cell Res. 118, 404-410.Mahowald, A. and Kambysellis (1980). Oogenesis. pp. 141-224. In,The Genetics and Biology of Drosophila, Vol. 2d, edited by M.Ashburner and T.R.F. Wright, Academic Press, London and New York.Marchant, G. and D.G. Holm (1988). Genetic analysis of theheterochromatin of chromosome 3 in Drosophila melanogaster. II.Vital loci identified through EMS mutagenesis. Genetics 120, 519-532.Maniatis, T., E.R. Fritsch, and J. Sambrook (1982). MolecularCloning, a Laboratory Manual. Cold Spring Harbor Laboratory, ColdSpring Harbor, New York.Mange, A.P., and L. Sandler (1973). A note on the maternal effectmutants daughterless and abnormal oocyte in Drosophilamelanogaster. Genetics 73, 73-86.McGhee, J.D., and G. Felsenfeld (1980). Nucleosome structure. Annu.Rev. Biochem 49, 1115-1156.Megee, P.C., B.A. Morgan, B.A. Mittman, and M.M. Smith (1990).Genetic analysis of histone H4: essential role of lysines subjectto reversible acetylation. Science 247, 841-845.183Michailidis, J., N.D. Murray, and J.A. Marshall Graves (1988) . Acorrelation between development time and variegated position-effectin Drosophila melanogaster. Genet. Res. 52, 119-123.Miklos, G.L.G., M.J. Healy, P. Pain, A.J. Howells, and R.J. Russell(1984). Molecular and genetic studies on the euchromatin-heterochromatin transition region of the X chromosome of Drosophilamelanogaster. I. A cloned entry point near the uncoordinated (unc)locus. Chromosoma 89, 218-227.Moore, G.D., J.D. Procunier, D.P. Cross, and T.A. Grigliatti(1979). Histone gene deficiencies and position-effect variegationin Drosophila melanogaster. Nature 282, 312-314.Moore, G.D., D.A. Sinclair, and T. Grigliatti (1983). Histone genemultiplicity and position-effect variegation in Drosophilamelanogaster. Genetics 105, 327-344.Mottus, R., R. Reeves, and T.A. Grigliatti (1980). Butyratesuppression of position-effect variegation in Drosophilamelanogaster. Mol. Gen. Genet. 178, 465-469.Muller, H.J. (1930). Types of variable variations induced by X-raysin Drosophila. J. Genet. 22, 299-334.Noble, M., S.A. Lewis, N.J. Cowan (1989). The microtubule bindingdomain of microtubule-associated protein MAP1B contains a repeatedsequence motif unrelated to that of MAP2 and Tau. J. Cell Biol.109, 3367-3376.Nurse, P. (1990). Universal control mechanism regulating the onsetof M-phase. Nature 344, 503-508.Nusslein-Volhard, C., E. Wieschaus, and H. Kluding (1984). Mutationaffecting the pattern of the larval cuticle in Drosophilamelanogaster, I. Zygotic loci on the second chromosome. WilhelmsRoux's Arch. Dev. Biol. 193, 267-282.O'Hare, K., and G.M. Rubin (1983). Structures of P transposableelements and their sites of insertion and excision in theDrosophila melanogaster genome. Cell 34, 25-35.Paro, R., and D.S. Hogness (1991). The polycomb protein shares ahomologous domain with a hterochromatin-associated protein ofDrosophila. Proc. Natl. Acad. Sci. USA 88, 263-267.Paro, R. (1990). Imprinting a determined state into the chromatinof Drosophila. Trends Genet. 6, 416-421.Peifer, M., F. Karch, and W. Bender (1987). The bithorax complex:control of segmental identity. Genes Dev. 1, 891-898.184Postlethwait, J.H., and K. Weiser (1973). Vitellogenesis induced byjuvenile hormone in the female sterile mutant apterous-four inDrosophila melanogaster. Nature New Biol. 244, 284-285.Postlethwait, J.H., A.M. Handler and P.W. Gray (1976). A geneticapproach to the study of juvenile hormone control of vitellogenesisin Drosophila melanogaster. pp. 449-469. In, The Juvenile Hormones,edited by L.I. Gilbert, Plenum Press, New York.Prokofyeva-Belgovskaya, A.A. (1941). Cytological properties ofinert regions and their bearing on the mechanics of mosaicism andchromosome rearrangement. Drosophila Inf. Serv. 15, 34-35.Prokofyeva-Belgovskaya, A.A. (1947). Heterochromatinization as achange of cell cycle. J. Genet. 48, 80-98.Reuter, G., R. Dorn, and H.J. Hoffmann (1982). Butyrate sensitivesuppressor of position-effect variegation mutations in Drosophilamelanogaster. Mol. Gen. Genet. 188, 480-485.Reuter, G., R. Dorn, G. Wustmann, B. Friede, and G. Rauh (1986).Third chromosome suppressor of position-effect variegation loci inDrosophila melanogaster. Mol. Gen. Genet. 202, 481-487.Reuter, G., J. Gausz, H. Gyurkoviks, B. Friede, R. Bang, A.Spierer, L.M.C. Hall, and P. Spierer (1987). Modifiers of position-effect variegation in the region from 86-88b of the Drosophilamelanogaster third chromosome. Mol. Gen. Genet. 210, 429-436.Reuter, G., M. Giarre, J. Farah, J. Gausz, A. Spierer, and P.Spierer (1990). Dependence of position-effect variegation inDrosophila on dose of a gene encoding an unusual zinc-fingerprotein. Nature 344, 243-244.Reuter, G., W. Werner, and H.J. Hoffmann (1982). Mutants affectingposition-effect heterochromatinization in Drosophila melanogaster.Chromosoma 85, 539-551.Reuter, G., and I. Wolff (1981). Isolation of dominant suppressormutations for position-effect variegation in Drosophilamelanogaster. Mol. Gen. Genet. 182, 516-519.Reuter, G., I. Wolff, and B. Friede (1985). Functional propertiesof the heterochromatic sequences inducing W 4 position-effectvariegation in Drosophila melanogaster. Chromosoma 93, 132-139.Roberts, P.A. (1965). Difference in the behaviour of eu- andheterochromatin: crossing-over. Nature 205, 725-726.Robertson, H.M., C.R. Preston, R.W. Phillis, D.M. Johnson-Schlitz,W.K. Benz, and W.R. Engels (1988). A stable genomic source of Pelement transposase in Drosophila melanogaster. Genetics 118, 461-185470.Rudkin, G.T. (1969). Non-replicating DNA in Drosophila. Genetics(Suppl.) 61, 227-238.Rushlow, C.A. and A. Chovnick (1984). Heterochromatic positioneffect at the rosy locus of Drosophila melanogaster: Cytological,genetic and biochemical characterization. Genetics 108, 589-602.Rushlow, C.A., W. Bender, and A. Chovnick (1984). Studies on themechanism of heterochromatic position effect at the rosy locus ofDrosophila melanogaster. Genetics 108, 603-615.Salas, F., and J.A. Lengyel (1984). New mutants. Drosophila Inform.Ser. 60, 243-244.Sandler,^L.^(1975).^Studies on the genetic control ofheterochromatin in Drosophila melanogaster. Israel J. Med. Sci. 11,1124-1134.Sandler, L. (1977). Evidence for a set of closely linked autosomalgenes that interact with sex chromosome heterochromatin inDrosophila melanogaster. Genetics 86, 567-582.Sandler, L., D.L. Lindsley, B. Nicoletti, and G. Trippa (1968).Mutants affecting meiosis in natural populations of Drosophilamelanogaster. Genetics 60, 525-558.Sawadogo, M. and M.W. Van Dyke (1991). A rapid method for thepurification of deprotected oligodeoxynucleotides. Nuc. Ac. Res.19, 674.Schalet, A. and G. Lefevre Jr. (1976). The proximal region of theX chromosome. pp. 847-902. In, The Genetics and Biology ofDrosophila, Vol. 1B, edited by M. Ashburner and E. Novitski,Academic Press, London and New York.Schultz, J. (1950). Interrelations of factors affectingheterochromatin-induced variegation in Drosophila. Genetics 35,134.Schultz, J. (1956). The relation of heterochromatic chromosomeregions to the nucleic acid content of the cell. Cold Spring HarborSymp. Quant. Biol. 21, 307-327.Shupbach, T., and E. Wieschaus (1986). Maternal-effect mutationsaltering the anterior-posterior pattern of the Drosophila embryo.Wilhelms Roux's Arch. Dev. Biol. 195, 302-317.Schupbach, T., and E. Wieschaus (1989). Female sterile mutations onthe second chromosome of Drosophila melanogaster. I. Maternaleffect mutations. Genetics 121, 101-117.186Schupbach, T., and E. Wieschaus (1991). Female sterile mutations onthe second chromosome of Drosophila melanogaster. II. Mutationsblocking oogenesis or altering egg morphology. Genetics 129, 1119-1136Sinclair, D.A.R., R.C. Mottus, and T.A. Grigliatti (1983). Geneswhich suppress position-effect variegation in Drosophilamelanogaster are clustered. Mol. Gen. Genet. 191, 326-333.Sinclair, D.A.R., Y.K. Lloyd, and T.A. Grigliatti (1989).Characterization of mutations that enhance position-effectvariegation in Drosophila melanogaster. Mol. Gen. Genet. 216, 328-333Sinclair, D.A.R., A.A. Ruddell, J.K. Brock, N.J. Clegg, V.K. Lloyd,and T.A. Grigliatti (1991). A cytogenetic and geneticcharacterization of a group of closely-linked second chromosomemutations that suppress position-effect variegation in Drosophilamelanogaster. Genetics 130, 333-344.Southern, E.M. (1975). Detection of specific sequences among DNAfragments separated by gel electrophoresis. J. Mol. Biol. 98, 503-517.Spofford, J.B. (1976). Position-effect variegation in Drosophila.pp. 955-1018. In, The Genetics and Biology of Drosophila, Vol. lc,edited by M. Ashburner and E. Novitski, Academic Press, New York.Spradling, A.C., and G.H. Karpen (1990). Sixty years of mystery.Genetics 126, 779-784.Spradling, A.C., and G.M. Rubin (1981). Drosophila genomeorganization: conserved and dynamic aspects. Annu. Rev. Genet. 15,219-264.Stern, C., and M. Kodani (1955). Studies on the position effect atthe cubitus interruptus locus of Drosophila melanogaster. Genetics40, 343-373.Strauss, F., and A. Varshaysky (1984). A protein binds to asatellite DNA repeat at three specific sites that would be broughtinto mutual proximity by DNA folding in the nucleosome. Cell 37,889-901.Szabad, J., G. Reuter, and M.B. Schroder (1988). The effects of twomutations connected with chromatin functions on female germ-linecells of Drosophila. Mol. Gen. Genet. 211, 56-62.Szidonya, J., and G. Reuter (1988). Cytogenetic analysis of theechnoid (ed), dumpy (dp) and clot (Cl) region in Drosophilamelanogaster. Genet. Res. Camb. 51, 197-208.187Tartof, K.D., C. Hobbs, and M. Jones (1984). A structural basis forvariegating position effects. Cell 37, 869-878.Turner, B.M. (1991). Histone acetylation and control of geneexpression. J. Cell Sci. 99, 13-20.Weisbrod, S., M. Groudine and H. Weintraub (1980). Interaction ofHMG 14 and 17 with actively transcribed genes. Cell 19, 289-301.Wilson, T.G. (1985). Determinants of oocyte degeneration inDrosophila melanogaster. J. Insect Physiol. 31, 109-117.Wohlwill, A.D., and J.J. Bonner (1991). Genetic analysis ofchromosome region 63 of Drosophila melanogaster. Genetics 128, 763-775.Wu, R.S., H.T. Panusz, C.L. Hatch and W.M. Bonner (1984). Histonesand their modifications. CRC Crit. Rev. Biochem. 20, 201-263.Wustmann, G., J. Szidonya, H. Taubert, and G. Reuter (1989). Thegenetics of position-effect variegation modifying loci inDrosophila melanogaster. Mol. Gen. Genet. 217, 520-527.Zhang, P., and R.S. Hawley (1990). The genetic analysis ofdistributive segregation in Drosophila melanogaster. II. Furthergenetic analysis of the nod locus. Genetics 125, 115-127.Zhimulev, I., E. Belyaeva, 0. Fomina, M. Protopopov, and V.Bolshakov (1986). Cytogenetic and molecular aspects of position-effect variegation in Drosophila melanogaster. Chromosoma 94, 492-504Zink, B., and R. Paro (1989). In vivo binding pattern of a trans-regulator of homeotic genes in Drosophila melanogaster. Nature 337,468-471.Zuckerkandl, E. (1974). Recherches sur les proprietes et l'activitebiologique de la chromatine. Biochimie 56, 937-954.188


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items