Copyright 0 1990 by the Genetics Society of America

The Molecular Analysisof the el-noc Complex of Drosophila melanogaster Terence Davis,Jennifer Trenear and Michael Ashburner Department of Genetics, University of Cambridge, Downing Street, Cambridge C82 3EH, England Manuscript received February 8, 1990 Accepted for publication May 14, 1990 ABSTRACT The el-noc complex spans a distance of about 200 kbon chromosome 215. It consistsof three discrete genetic regions el, 1(2)35Ba and noc, each of which has a distinct phenotype when mutant. The noc locus itself is complex, including three separate regions. The el locus has been characterized by mapping 30 aberration breakpoints to the DNA. It extends over a distance of about 80 kb. It can be divided into two parts by the aberrations In(ZLR)DTD128 and T(Y;Z)A80.These break between two sets of el alleles yet are both phenotypically wild type for elbow. The simplest explanation is that el consists of two transcription units elA and elB. The locus pu, which appears to be unrelated to the el-noc complex, is found to map between the two el loci very close to elB (the distal el locus). The loci 1(2)35Ba and nocA have been separated by only two 1(2)35Ba+nocA- deletions and a nocA- inversion. No 1(2)35Ba-nocAC aberrations have been found. At the molecular levelthese loci are found to occupy ~, almost the same region, and are probably identical1.

T

HE el-noc region has been extensively described (WOODRUFF and ASHBURNER1979a;ASHBURNER,AARONand TSUBOTA 1982; ASHBURNER, TSUBOTA and WOODRUFF 1982) andincludes at least three distinct genetic loci, elbow ( e l ) , 1(2)35Ba and no-ocelli (noc), all separable by chromosomal aberration breakpoints. Mutations of el show a wing and haltere phenotype, mutations of 1(2)35Ba are larval lethals and mutations of noc lack the ocelli. All three of these loci show partial failure of complementation. Flies heterozygous for 1(2)35Ba and noc or el alleles frequently have weak phenotypescharacteristic of either,or both, of these loci. Alleles of noc heterozygous with some el alleles have weak no-ocelli phenotypes. 1(2)35Ba homozygotes are usually lethal, but some, e.g., 1(2)35BaAR”,are hypomorphic lethals and can survive at low frequency when heterozygous with a 1(2)35Ba- deletion. The escapers usually have a weak no-ocelli phenotype, if the deletion is also el- then they also have a weak elbow phenotype. Flies which are heterozygous for different mutant1(2)35Ba alleles may survive at low frequency and have a weak elbow and a weak no-ocelli phenotype. These threeloci also interact with thedominantantimorphicmutation Tp(2;Z)Scutoid (Sco), an X-ray-induced mutation (KRIVSHENKO1959) thatis associated with a reciprocal transposition between noc, osp (outsread)and Adh (Alcohol dehydrogenase) and six loci from 35D (ASHBURNER, TSUBOTA and WOODRUFF 1982; MCGILLet al. 1988). The Scutoid phenotypeis a loss of bristles from the head and thorax of the adult fly. The wild-type number of these bristles is 40, whereas in Sco/+ flies the number is reduced to between 25 and 29 (ASHBURNER, TSUBOTA and WOODRUFF1982). Sco is a Genetics 1 2 6 105-1 19 (September, 1990)

recessive lethal, although very occasional Sco homozygotes (1 1/13,475)survive to adulthood. These flies have a very extreme Scutoidphenotype with only eight to ten bristles per fly. Some alleles of el, all 1(2)35Ba alleles and some noc alleles act as dominant enhancers of the Scutoid phenotype, when heterozygous with the Sco chromosome they reduce the number of bristles by between 5 and 15 perfly. When Sco is heterozygous over deletions that include el, 1(2)35Ba or noc, a progressive enhancement of the Scutoid phenotype is observed which is proportional to the number of these loci deleted. T h e maximal enhancement is observed when all three are deleted; these flies have a similar phenotype to Sco homozygotes. Duplications of the region that include e l , 1(2)35Ba and noc, e.g., Dp(2;2)Adh3,partially suppress the Scutoid phenotype, increasing the numberof bristles to about 34 per fly (ASHBURNER, TSUBOTA and WOODRUFF 1982). These datasuggest that el, 1(2)35Ba and noc comprise a genetic complex. T h e proximal part of this complex, including the noc locus, was cloned by CHIAet al. (1985a). In this study the cloning of the remaining part of the complex is described. T h e relationships of the loci comprising the complex are discussed both with respect to each other and to the neighboring loci. T h e genetic relationships of these loci, based upon chromosomal aberrationsbroken within this region, are shown in Figure 1. MATERIALS AND METHODS Drosophila stocks: The chromosomes used in this study are listed in Table 1. The A series of deletions were X-rayinduced Adh null mutations (AARON1979; ASHBURNER,

106

T. Davis, J. Trenear and M. Ashburner TE146(Z)

TE146(Z)T(Y;2)GR18

T(2;3)GTIO T(2;4)GT6

-"" 7p

V)34Fb

1(2)34Fc

1I 1I 1

1(2)34Fd mr(2)34Fe 1(2)35Aa

elB

-I L I I I I I 1 1 1 1 PU

elA

1(2)35Ba

nod

FIGURE 1.-Genetic map of the region surrounding el on chromosome 2L of D. melanogaster. T h e genetic loci are 1(2)34Fb, l(2)34Fc, 1(2)34Fd, ms(2)34Fe, 1(2)35Aa, elbow B ( d B ) ,pupal ( p u ) ,elbow A (elA), l(2)?5Baand no-ocelli A (nocA). T h e genetic breakpoints of deletions, inversions and translocations are shown; all aberrations are deletions unless otherwise indicated. Those breakpoints located on the DNA cloned are indicated by double vertical lines.

AARONand TSUBOTA 1982). Thefn series were formaldeet al. 1977; hyde-induced Adh null mutations (O'DONNELL ASHBURNER, AARON and TSUBOTA 1982). The derivatives of TE146 are described by GUBBet al. (1985, 1986) and those of TE36 by GUBBet al. (1984). The el' mutation was spontaneous (BRIDGES and BREHME1944), Df(ZL)683d29a is neutron-induced a deletion (ALEXANDROV 1986). In(ZLR)DTD128 and T(2;3)shv19 were induced by W. GELBART (SMOLIK-UTLAUT and GELBART 1987; SEGAL and GELBART 1985). T(Y;Z)A80 was induced by LINDSLEY et al. (1972). The translocation T(Y;Z)A15 was induced by T. LYTTLE(unpublished results). All other aberrations were induced in Cambridge. Flies were grown on either yeast/ glucose medium, Phillip-Harris Instant Drosophila food or cornmeal agar. Preparation of DNA from Drosophila: DNA was prepared from adult flies as described by CHIAet al. (1985a). Preparationof DNA from phage:Late log phase cultures of Escherichia coli K802 grown in 10 g/liter Bacto-tryptone, 5 g/liter yeast extract, 5 g/Iiter NaCI, 10 mM-MgnS04 were infected with phage at a multiplicity of infection of either 1 (Charon 4 phage) or 0.1 (EMBL3 or EMBL4 phage). Growth was allowed to proceed for 1 to 4 hr at 37" with aeration until cell lysis was apparent. Cellular debris was removed by centrifugation at 5000 X g for 10 min. The phage in the supernatant were concentrated by the addition of polyethyleneglycol 6000and NaCl as described by MANIATIS, FRITSCHand SAMBROOK (1982) and were further purified on a CsCl step gradient with step densities of 1.3, 1.5 and 1.7 g/mlfollowed by centrifugation on CsCl equilibrium gradient. The purified phage bands were dialyzed against 10 mM Tris-HCI (pH 8.0), 1 mM-EDTA and extracted with phenol and thenwith chloroform. The DNA in the aqueous

phase was precipitated with 2 volumes of ethanol, washed with 70% ethanol and resuspended in 10 mM Tris-HCI (pH 8.0), 1 mM EDTA. Preparation of probes: Probes were prepared as descibed in CHIAet al. (1 985a). Chromosome walking, screening and restriction mapping: Chromosome walking was performed essentially as described by BENDER,SPIERERand SPIERER(1983)and et al. (1983). In this study three libraries have been BENDER used, the LAUER-MANIATIS library (MANIATIS et al. 1978), a library made from Canton S DNA in EMBL4 phage (proand a library constructed by J. vided by M. GOLDBERG) TAMKUN from a stock made by J. KENNISONcarrying the markers y; cn bw sp and isogenic for all four chromosomes personal communication). (in EMBL.3 phage) (J. TAMKUN, Phage were mapped with EcoRI, HindIII, Sal1 and XhoI by running all possible single and double digests on a single agarose gel. Mapping breakpointsby Southern hybridization:Nicktranslated clones were hybridized to enzyme-digested DNA prepared from flies bearing chromosomes that were suspected to have their breakpoints lying within the cloned region of DNA. All lethal chromosomes used were balanced over In(ZLR)O, Cy dp'" AdhnBpr cn2 (Cy0ne)or In(ZLR)O, Cy dp'"' p r cn2 (CyO) unless otherwise stated. Parental chromosomes were homozygous except T E 3 6 ( R ) and TE146(Z), which carry lethal mutations. Since many of the aberrations that have been analyzed were induced onknown progenitor chromosomes (Table l), direct comparison of the aberrations with their precursors was usually possible. The Southern hybridizations were performed as described in CHIAet al. (1985a). Restriction enzyme digests of genomic DNA were with EcoRI unless otherwise stated.

of

Analysis

the el-noc Complex

107

TABLE I Chromosomes used in this study, grouped according to their progenitor chromosomes Aberration

(1) On b Adh'cn bw Df(ZL)A245 Df(2L)A26U Df(2L)A266 Df(2L)AMU In(ZLR)e/6 (2) On Adhuf3cn Df(2L)elI 4 Df (2L)el I6 ln(2L)elY T(2;3)el24 (3) On AdVf3p r cn Df(2L)el28 (4) On b Adh"" pr cn bw T(Y;2)el4 (5) On a1 d p b TE146(Z) pr l(2)pwn cn Df(ZL)TEl46(Z)GWI Df(ZL)TE146(Z)GW6 Df(2L)TE146(Z)GW7 Df(2L)TE146(Z)GW12 In(2L)TE146(Z)GW13+ Df(2L)TE146(Z)GW13 T(Y;2)TEl46(Z)GRl8"Rl5" (6) O n b ~ k ' ' ' ~ " ~pr) pk cn sp Df(2L)TE36(R)GW8 (7) O n Adh" pr cn Df (2L)fnZ Of(2L)fnj (8) On Adh' p r pk cn sf Df (2L)b84a2 (9) On a known wild-type chromosome 2 T(Y;2)GT2 T(2;3)GT3 (IO) On a known wild-type chromosome 2 T(2;4)GT6 T(2;3)GT8 T(2;3)GTIU (1 1) N o progenitor chromosomes available Df(2L)ARRl Df(ZL)b83d29a + In(2L)b83d29a Df (2L)elI In(ZL)b83b22 In(ZLR)DTD128 T(Y;2)A15 T(Y;2)A8U T(2;3)shv19 . .

Mapping breakpoints by hybridization in situ: Stocks in which the aberrationof interest was balanced over asuitable balancer [usually Cy0 or In(2LR)GlaI were used for the mapping of the clones to polytene chromosomes. Salivary gland polytene chromosomes were prepared andhybridized with labeled probes as described by GUBBet al. (1984). Probes were labeled by nick-translation using DNA polymerase I according to the procedure of GUBBet al. (1984) using Biotin-16-UTP (Boehringer) at a final concentration of 0.1 pmol per ml in the presence of 30 p~ each of dATP, dGTP and dCTPfor 90 min at 16". After hybridization to polytene chromosomes the hybridized DNA was visualized by staining with horseradish peroxidase. Chromosomes were washed twice in phosphate buffered saline (PBS) at room temperature for 5 min, followed by PBS containing 0.1 %

Cytology

Df(2L)35A4;35B2 Df(ZL)35B1.2 Df(ZL)35B2.3 Df(2L)35A1-4;35BlO ln(2LR)35B1.3;57C3-9 Not cytologically visible Not cytologically visible In(ZL)34A2.3;35A3.4 T(2;3)35B1.3;93C3-7

Df(ZL)35B3;35D4 T(Y;2)Y;35B1 Df(2L)35B1;35C4.5 Df(2L)35B1;35C1 Df(2L)35A3.4;35B2 Not cytologically visible ln(2L)35B1.2;40 + not cytologically visible Df(2L)35B1.2;35B8.9-C1 Df(2L)35B1.2;35E1.2 Df(ZL)35A3;35BZ Df(ZL)35B1;35B3.4 Df(2L)34C3;35A4 from the stock L w'lE4 T(Y;Z)Y;35A1.2

T(2;3)28BC;35Bl.2;41;42F;98D from the stock z w'"' T(2;4)34F3;101-102 T(2;3)35A1-4;62F3-6; + T(2;3)50B;81 T(2;3)35A1-4;76A5-7

Df(2L)35A4-B1.2;35B8.9-C1

Df(2L)34C7.8;35A-B + B;35D5-7 Not cytologically visible In(ZL)34D4.5;35C1.2 In(ZLR)35B1.3;48C6-8 T(Y;2)BSYy+;35A4-B1.2 T(Y;2)YS;35A3.4 T(2;3)22F1.2;35A1-4;97A .

In(2L)35A-

.

Triton X-100 for 2 min and rinsing in PBS. Chromosomes were then incubated in the presence of a 1:250 dilution of streptavidin-biotinylated peroxidase complex (Uniscience, Ltd.) for 30 min at 37". After incubation the above wash steps were repeated. Chromosomes were then incubated at room temperature for 1-10 min in the presence of 0.5 mg per ml 3,3'-diaminobenzidine (Sigma) in PBS containing 0.01% H202. The staining reaction was stopped by immersion in water followed by immersion in PBS. Chromosomes were then stained with 5% Giemsa (GUBBet al. 1984). RESULTS

The map: This study is an extension of the work started by CHIAet al. (1985a) andMCGILLet al. ( 1 988)

T. Davis, J.and Trenear

108

I1

I I I

I

I

I

II

II

I

1

I

I

~

-ob26.04 -ob%.lQ -0b26.18 “-----ob27.06 -0bZ7.03 ob28. 13

11

I

I

sb18.41 obl7.41 ob18.42

l

l

Xho I

-0blO.02 ob12.07

ob19.03 Ob19.48

~

dsc35

-ob25 13

IP

I I

xhal

ob25.02

-0b30.05 ob30.012 Db30.03

-

I1

-ob23.05 Ob24.05 -obll.O2 -ob25.2

ob28.23 -0b28.19

-ob29.08 ~

I

M. Ashburner

ob13.01 ob14.21 ob14.04

Ob20.43

-Ob20.32 ob20.13 ob21.11 ob22.35

ob15.85

-

Obl6.12 ob16. 19 ob16.01 obi6.22 obi7.36

-

-ob22.29 FIGURE2.-Map of 250 kb of DNA from the el region of D.melanogaster. Restriction enzyme sites for EcoRI, HindIII, Sal1 and XhoI are s11ow11lxlow the coordinates. T h e map is a composite of DNA from phage made using DNA from Canton S flies cloned into Charon 4 , I:“BL? and EMBL.4 phage (see MATERIALS A N D METHODS). Only representative phage are shown. T h e origin of the map coordinates is the I:’coKI site 1321 bp distal t o the start of transcription of the proximal A d h mRNA (BENYAJATI, PLACEand SOFER1983; CHIAet al. 1985a) and are negative i n the direction of the telomere. P indicates a polymorphic site. T h e Sal1 sites distal to the phage Xob25.13 have not been ~napped.

involving the molecular analysis of the Adh region of the genome of Drosophilamelanogaster. CHIAet al. (1985a) cloned the genetic loci Adh, osp and noc in overlapping lambda phage by chromosomal walking. This walk finished 124 kb distal to the Adh locus with the phage Xob10.02. In this study this walk has now been extended to 360 kb distal to Adh (Figure2). Two strategies were employed: (1) the walkwas extended from Xob10.02 by chromosomal walking in the distal direction; (2) a bidirectional walk was initiated fromthephage Xdsc35 (provided by D. ANDREWS) which was thought to map to one of the loci immediately distal to the el-noc complex. The phage Xdsc35 maps at -286.5 to -298.0 (Figure 2). The map coordinates are negative in the distal direction. All phage between Xob10.02and Xb30.012 inclusive have been checked by hybridization to polytene chromosomes of a wild-type stock (Canton S) and cover the region 35A1.2 to 35B1.2. The phage Xob18.42 has a second minor site at the polytene chromosome region 93E1.2 of 3R andthe phage Xdsc35 has a second site at 57El-4 of 2R. The positions of all breakpoints which are notdeletions of genomic DNA have been confirmed by in situ hybridization of polytene chromosomes with a suitable probe. The -1 10.0- to -150.0-kb region: This region is covered by the phage Xob10.02 to Xobl4.04 (Figure 2) and includes the distal part of nocA, 1(2)35Ba and the proximal part of the el locus. The proximal limit for the locus 1(2)35Ba is defined by the 1(2)35Ba+ nocdeletionDf(2L)A446 at coordinates -1 13.5 to -1 17.5 (CHIAet al. 1985a)andthe most proximal of the 1(2)35Ba- noc- deletions Df(2L)Adhn78‘3at -1 19.5 to -121.8 (MCGILL1985). The distal limit of 1(2)35Ba is defined by the el- 1(2)35Ba+deletion Df(2L)b84a2. When DNA from Df(ZL)b84~2/CyOdigested with

EcoRI is probed with the 4.6-kb EcoRI fragment of Xobll.02 (Figure 4) two fragments are seen, the 4.6kb parental fragment and a fusion fragment of 3.3 kb (Figure 3, lane 4). This is not seen in the parental DNA (Figure 3, lane 5). When Df(2L)b84a2 is heterozygous over Df(ZLyn3 (an el+ 1(2)35Ba- deletion) the 4.6-kb fragmentis absent but the 3.3-kb fragment is present. In addition a second fragment of 1.5 kb is found (Figure 3, lane 3). This is also present in the Df(2Lvn3/CyO lane(Figure3,lane 2) and is thus the fusion fragment of Df(ZLyn3. Therefore both Df(2L)b84a2 and D f ( 2 L y n 3 are broken in the 4.6-kb EcoRI fragment at -121.8 to -125.1 and -124.9 to -126.4, respectively. Four other el+ 1(2)35Ba- aberrations are broken within this region, Df(2L)A245, Df(2L)A260,Df(2L)A266 the and synthetic deletion T(Y;2)TE146(Z)GR18”R15P.Df(2L)A245/CyO nB EcoRIdigested DNA probed with the 7.2-kb EcoRI fragment of Xob12.07 (Figure 2) gives the 7.2-kb parental and a 3.2-kb fusion fragments (Figure 3, lane 7). This deletion is broken between -134.6 and -137.8. DNA from Df(2L)A260 and Df(2L)A266 [both heterozygous with Df(2L)b84a2] probed with the 1.9-kb and 2.4-kb EcoRI fragments of Xob14.21 (Figure 5) show only the 2.4-kb fragment (Figure 3, lanes 9 and 10). There are no fusion fragments visible, therefore these deletions are brokenclose to theEcoRI site at - 144.4. These deletions include the genetic loci 1(2)35Ba,noc, osp and Adh but donot include1(2)35Bbthe nextmost proximal locus. As the distal endpoints of these deletions are indistinguishable, it is possible that they are re-isolates of the same event. DNA from flies heterozygous for T(Y;2)TE146(Z)GR18DRl5‘and Df(2L)b84a2 lacks the 2.8kb, 1.5-kb and 7.2-kb fragments when probed with Xob12.07 but retains the distal 3.9-kb fragment (Fig-

Analysis of the el-noc Complex 1 23 4

5

6

8 91 1 0

7

109

11321 4116151871 9

-

"

7.2

4.

4.0 2.7

3.9

2.4

2.8

'.5 L,

3'J'

--

1.5

(a)

(b)

(c)

(dl

(e)

(f)

(9)

FIGURE3.-Southern transfers of EcoRI-digested genomic DNA probed with purified DNA fragments or X phage covering the region of the WIIL between - 1 10.0 and - 1 50.0. Thesizes of wild-type restriction enzyme fragments are indicated; f indicates a fusion fragment. The EcoRI site at -141.7 is present on the a1 d p b TE146(Z)pr 42)pwn cn and the b Adh' cn bw chromosomes and their derivatives and also on the In(2LR)O.Cy dp"" Adh"" pr cnz chromosome. The bands in tracks 16 and 17 marked X are at SA1 on the X chromosome (CHIAel a/. 1985a). (a) Probed with the 4.6-kb EcoRl fragment of Xob11.02: Adh" pr cn (track 1); Df(2Llfn3, pr cn/CyO (track 2); Df(2Llfn3, p r cn/ DJ(21.)b84a2,Adh' p r p k cn 5 p (track 3); Df(2L)bR4a2,Adh' pr p k cn sp/CyO (track 4); Adh' pr p k cn sp (track 5); (b)probed with the 7.2-kb EcoRI fragment of Xob12.07 b Adh' cn bw (track 6); Df(2L)A245,b cn bw/CyO nB (track 7); (c) probed with the 1.9-kb and 2.4-kb EcoRI fragments of Xob14.21:b Adh' cn bw (track 8); Df(2L)A260,b cn bw/Df(2L)b84a2, A d h F p r p k c nsp (track 9); Df(2L)A266,b cn bw/Df(2L)bR4~2, Adh' pr p k cn 5p (track 10); (d) probed with Xob12.07:a1 d p b TE146(Z) p r l(2)pwn cn/CyO (track 1 1); T(Y;2)TE146(Z)GRIR"RI5', a1 d p b/ Df(2L)bX4a2.Adh'pr p k cn 5p (track 12): (e)probed with the 4.0-kbEcoRI fragment of Xob14.04:a1 d p b TE146(Z)pr 1(2)pwncn/C@ (track 13); Df(2L)TE146(Z)CW1, a1 d p b pr l(2)pwn cn/CyO (track 14); Df(2L)TEI46(Z)GlV6, a1 d p b pr l(2)pwn cn/CyO (track 15); (0 probed with the 8.5-kb BcoKI fragment of Xob10.02:L wl'"; (track 16); L w'"'; T(2;3)CTR/CyO(track 17); (g) probed with the 5.4-kb Hind111 fraglnent o f Xob11.02:Xhol-digested Adh"f3 cn DNA (track 18); Xhol-digested T(2;3)e124. Adh"f3cn/CyO DNA (track 19).

+

ure 3, lane 12). There is a faint fragment at about 1.5 kb which is presumably the fusion fragment. As the 3.9-kb fragment is due to a polymorphic EcoRI site at -141.7 (Figures 2 and 5) the chromosomal 2 breakpoint of T(Y;2)GR18" must be close to the EcoRI site at -1 37.8. Two el- l(2)35Ba- deletions, Df(2L)TEI46(Z)GWI and Df(2L)TE146(Z)GW6 are broken in the 4.0-kb EcoRI fragment of Xob14.21 (Figure 5) and have fusion fragments of 4.5 kb and 5.2 kb, respectively (Figure 3, lanes 14 and 15). T h e fusion fragment of Df(2L)TE146(Z)GW6 gives a slightly weaker band of hybridization than that of Df(2L)TE146(Z)GWI, suggesting thatthe distal breakpoint of Df(2L)that of TE146(Z)GW6 is more distal than Df(2L)TE146(Z)GWI.T h e distal limits of these deletions are between -146.8 and -150.8. T h e 1(2)35Ba- not- translocation T(2;3)GT8probed with the 8.5-kb EcoRI fragment of Xob10.02 (Figure 4) has the parental fragment of 13.0 kb and fusion fragments of 1.5 kb and 20.0 kb (Figure 3, lane 17). This aberration is therefore broken within 1.5 kb of the EcoRI site at -1 19.5. T h e significance of this translocation is discussed later. Two el- alleles also have their breakpoints in this

region. T h e spontaneous allele el' is homozygous viable and has astrong elbow phenotype. T h e DNA from el' completely lacks homology with Xob14.21 (data not shown). Thus the el' chromosome carries a deletion in this region. T h e proximal end of the el' deletion is found in the 6.4-kb EcoRI fragment of Xobl3.01 (Figure 5). When el' DNA is probed with this fragment,a 3.9-kb EcoRI fragment is present (Figure 6, lane 1). T h e 3.9-kb fragment is due to a polymorphic site at -141.7 (Figure 5). This DNA also lacks the 5.6-kb HindIIIfragmentbutretainsthe adjacent 5.8-kb fragment(Figure6,lane 2). With XhoI only a fusion fragment of 19.0 k b is seen (Figure 6 , lane 3). Therefore el' is broken in the 1 1.5-kb XhoI fragment close to theEcoRI site at -141.7. DNA from T(2;3)el24/CyO probed with the 5.4-kb HindIII fragment of Xobll.02 (Figure 4) has fusion fragments in a XhoI digest of 8.4 kb and 13.0 kb (Figure 3, lane 19). T h e parental XhoI fragment is 6.5 kb. This translocation is broken between - 126.0 and -1 3 1.4. The -150.0- to -200.0-kb region: This region is covered by the phage Xob15.85 to Xob17.41 (Figure 2) and includes most of the el locus and the proximal part of the pu locus. Four el- l(2)35Ba- deletions that are wild type for pu have been analyzed, Df(2L)el28,

T. Davis, J. Trenear and M. Ashburner

110

E

E

I I

I

2.8

I

5.2

H

E

E 1.5

I H

I

6.5

X

E

E

4.6

1

2.3

I

13.0

4.6

I

2.3

I

1.4/ob11.02

I

2.3

I

I/

FIGURE4.-Expanded map of the region covering the phage Xob10.02 and Xobll.02. The coordinates are shown above the EcoRI (E) sites. Only the relevant Hind111 (H) and XhoI (X) sites are shown.

1

X 2.8

I 1.5 I

2.2

8.5

ob10.02

P

I

D

8.7 12.5

I

13.5

P

-

I

157

I

11.5

s

lfi R

3.4

I

3.6

1.6

II

4.0

I

2.4

I

II

4.0

I

2.4

I

1.9

ob1421 6.4

ob13.01

FIGURE5.-Expanded map of the region covering the phage Xob13.01, Xob14.21 and Xob15.85. The coordinates are shown above the EcoRI sites. Sites for HindIII (H), Sal1 (S) and XhoI (X) are also shown.

Df(2L)TE36(R)GW8,Df(2L)eZI4 and Df(2L)e116. DNA from Df (2L)eZ28/Df (2L)b84a2 heterozygotes when probed with Xob14.21 (Figure 5) lacks the 6.4kb, 2.4-kb and 4.0-kb fragments but retains the 3.6kb and 18.0-kb fragments (Figure7, lane 2). The 0.5kb fragment is not seen and there is no fusion fragment, therefore the breakpoint of Df(2L)e128 is between -150.8 and -151.3. DNA from Df(ZL)TE36(R)GW8/CyOhas been digested with Sal1 and XhoI and probed with the 8.7-kb XhoI fragment of Xob15.85 (Figure 5). T h e parental strain has 26.0-kb, 13.5-kb and 12.5-kb Sal1 fragments (Figure 7, lane 3). The 13.5-kb and 12.5-kb fragments are due to a polymorphic site on the Cy0 nB chromosome within the 26.0-kb fragment (Figure 5).The Df(2L)TE36(R)GW8chromosome lacks the 26.0 kb and has a 9.0-kb fusion fragment (Figure 7, lane 4). This chromosome also has a 7.5-kb XhoI fusion fragment (Figure 7, lane 6). The deletion is broken between -166. l and the XhoI site at -170. l . Df(2L)ellGICyO gives a 7.8-kb fusion fragment when probed with the 5.4-kb EcoRI fragment of Xobl7.41 (Figure Sa) and is therefore brokenbetween -187.9 and -193.3(Figure 7, lane 8). HindIII-digested DNA from Df(2L)eZI4/CyOgives a fusion fragment of 5.8-kb (Figure 7, lane 12)when probed with the 4.2-kb Hind111 fragment of Xob18.42 (Figure Sa). This deletion is broken between -196.0 and -200.2. The neutron-induced deletion Df (2L)b83d29a in-

cludes pu and atleast 18 moredistal loci. The deletion is associated with a small inversion between 35A-B (proximal to pa)and 35D5-7. As there is no parental chromosome available for this deletion, the inversion Zn(2L)b83b22,which was induced on the same progenitorchromosome, has been used as acontrol. When DNA from Df(2L)b83d29a/Zn(2LR)SM5, a12 Cy It" cn2 sp2 is probed with the 9.6-kb EcoRI fragment of Xob17.36 (Figure 8b) afusion fragment of 7.4 kb is seen (Figure 7, lane 13). The 9.6-kb fragment is due to a polymorphic EcoRI site in the 16.0-kb EcoRI fragment at -178.3 (Figure 8b). With DNA cut with XhoI a fusion fragment of 16.0 kb is seen (Figure 7, lane 14). Thesearenotfound with Zn(2L)b83b22 (Figure 7, lanes 15 and 16) therefore this aberration breaks between -178.3 and -185.7. T h e fusion fragments are from the 35A-B;35D5-7 inversion. Four other breakpoints are located in this region, two of which [el' and Zn(2LR)el6]are associated with el- alleles. T h e distal breakpoint of el' is broken in the 8.7-kb XhoI fragment of Xob15.85 (Figure5). When DNA from el' is probed with this fragment, the only fragments seen are the 8.7-kb EcoRI (due to a polymorphic EcoRI site at -164.4, see Figure 5) and 20.7-kb HindIIIparentalfragmentsanda 19.0-kb XhoI fusion fragment (Figure 6,lanes 4 , 5 and 6). T h e 19.0-kb XhoI fragment is the fusion fragment seen with the 6.4-kb EcoRI fragment of Xobl3.01 (Figure 6, lane 3). Therefore el' is broken close to the HindIII

Analysis of the el-noc Complex 1 2 3

4

5 6

20.7

t t

1

5.8

8.5

--

3.9

FIGURE6.--Southern transfers of genomic DNA fromDj(2L)el'. T h e sizes i n kt> of wild-typerestrictionenzymefragmentsare intlic;ltetl; f intlic;ltes ;I fusion fragment. T h e el' chronlosome has an EcoRI site at -141.7. ( a ) Probed with the 6.4-kb EcoRl fragment o f Xobl3.1)1: EcoRI-digested DNA (track 1); Hindlll-digested DNA (track 2): Xhol-digested DNA (track 3); (b) probed with the 8.7-kb Xhol fragment of Xohl5.85 (tracks 4, 5 and 6 as for tracks 1, 2 and 3).

sites at -1 62.2 to -163.1 and is a deletion of about 21.4 kb. In(2LR)elblCyO has two fusion fragments when probed with the 5.4-kb EcoRI fragment of Xob17.41 (Figure sa), at 2.3 kb and 7 kb (Figure 7, lane lo), so In(2LR)el6 breaks between -187.9 and -193.3. In(2LR)DTD128 is a y-ray-induced inversion (SMOLIK-UTLAUT and GELBART 1987), the left hand breakpoint of which was mapped genetically between pu and el. By in situ hybridization to polytene chromosomes the inversion break was found to map within the phage Xob16.19 (Figure1lb). As there is no progenitor chromosome available, DNA from In(2LR)DTD128/CyOhas been digested with four different restriction enzymes. EcoRI, HindIII, SalI and XhoI-digested DNA was probed with the 5.2-kb EcoRI fragment of Xob16.19 (Figure8b). With EcoRI the parental fragment of 8.5 kb (due to a polymorphic site) and fusion fragments of 20.7 kb and 1.9 kb are seen (Figure 9, lane 1). With HindIII only one fusion is seen at 18.5 kb (Figure9, lane 2). With SalI-digested DNA the 12.5-kb parental and fusion fragments of 28.0 kb and 24.0 kb are found (Figure 9, lane 3). XhoI-digested DNA gives the 10.9-kb parental fragment (due to a polymorphic XhoI site at -171.6) and fusions of 20.0 kb and 3.4 kb (Figure 9, lane 4). This inversion is therefore broken between -1 67.1 and -1 72.9.

111

The other el+ defining breakpoint, T(Y;2)A80, induced by LINDSLEY et al. (1 972),was also genetically mapped between pu and el. DNA from T(Y;2)A80/ Df(2L)b84a2 probed with Xob16.19 (Figure 8b) lacks the 8.5-kb fragment but retains the 5.4- and 9.5-kb fragments (due toa polymorphic site at -1 78.3 in the 16.0-kb fragment, Figure 8b). T h e 16.0-kb fragment is also present in this DNA therefore some of the chromosomes lack this polymorphic site. Fusion fragments of 19.0 and 20.0 kb are present (Figure 9, lane 5). With HindIII-digested DNA the parental 20.7-kb fragment is missing but fusion fragments of 1 1.5 and 25.0 kb are found (Figure 9, lane 6). Thus the translocation is broken between - 167.1 and -1 72.9. This has been confirmed by in situ hybridization to T(Y;2)A80 polytene chromosomes using Xobl6.19 as a probe (Figure 1 le). T h e deficiency Df(2L)ARRI is an X-ray-induced reconstitution of chromosome 2 from the y2" element of the translocation T(Y;2)A80 and the 2"Y" element of the translocation T(Y;2)R15 (WOODRUFF and ASHBURNER1979a;GUBB et al. 1984). Thus it has only the distal element of the translocation T(Y;2)A80. When DNA from Df(2L)ARRIICyO is probed with Xob16.19 the EcoRI fusion fragment of 20.0 kb and the HindIII fusion fragment of 25.0 kb are seen (Figure 9, lanes 7 and 8). This confirms that T(Y;2)A80 breaks between -167.1 and -172.9. The -200.0- to -260.0-kb region: This region is covered by the phage Xob18.42 to Xob22.36 (Figure 2) and includes the distal part of the el locus, the pu locus and the proximal part of the locus 1(2)35Aa. Four elaberrations T(Y;2)el4, T(Y;2)A15, T(2;3)shv19 and In(2L)el9 are broken in this region. DNA from flies heterozygousfor T(Y;2)el4 with Df(2L)b84a2 when probed with Xob18.42 (Figure 8a)lacks the 2.7-kb and 0.5-kb fragments but retains the 1.9-kb, 5.2-kb, 2.5kb, 1.7-kb and the 18.0-kb fragments (Figure 10, lane 2). When probed with only the 2.7-kb fragment it lacks this fragment and nofusion fragments are present (data not shown). When Xob18.42 is hybridized to polytene chromosomes of T(Y;2)el4 the T(Y;2) break is seen (Figure1 la). Thus the distal translocation breakpoint of T(Y;Z)el4 is between -207.3 and -207.8, and the proximal translocation breakpoint is close to the EcoRI site at -204.6. This event is associated with a small deletion of about 2.7 kb. DNA from T(Y;2)A15/Df(2L)b84a2 was digested with EcoRI and SalI and probed with the 8.5-kb XbaI fragment of Xob18.42 (Figure 8a). With EcoRI the 2.7-kb fragment is missing but the 2.5-kb, 1.7-kb and 18.0-kb fragments are retained (Figure 10, lane 3). In addition a single fusion fragment of 27.0 kb is found. With SalI the 8.1-kb fragment is missing and a fusion fragment of 27.0 k b is found (Figure 10, lane4). As only one fusion fragment is found with each enzyme, it is

T. Davis, J. Trenear and M. Ashburner

112 1

2 3

4

5

6

7

8

194 111 503 1 61 2 -

11 .

.).

18.(

6.d

4.c

3.f 2.d

FIGURE7,"Southern transfers of restriction enzyme-digested genomic DNA probed with purified DNA fragments or phage covering the region between - I 50.0 and -200.0. The sizes in kb of wild-type restriction enzyme fragments are indicated; f indicates a fusion fragment. I IIC Sal1 site at -163.4 is present on the In(2LR)O. Cy dp"" p r en2 chronlosome. (a)EcoRI-digested DNA probed with Xob14.04: Adh""pr cn (track 1); Df(2L)tl28,pr cn/Df(2L)b84a2, Adh'pr p k cn sp (track 2): (b)DNA probed with the 8.7-kbXhol fragment of Xob15.85: Sall-digested b ck'rmw) pr p k en sp/CyO nR D N A (track 3): Sall-digested Df(2L)TE36(R)C11'8, bpr p k en sp/C;jO n R D N A (track 4): Xhol-digested b ck" '"'")pr r .

p k cn s p / C y O nB DNA (track 5); XhoI-digested Df(2L)TE36(R)Gll4?, bpr p k en sp/CyO nR DNA (track 6); (c) EcoRI-digested DNA probed w i t h the 5.4-kb BcoKI fragment of Xobl8.41: Adh"/' en (track 7); Df(2L)el16, cn/CyO (track 8); (d) EcoRI-digested D N A probed with the 5.4kb EmRI fragment of Xoh18.41: b Adh' en bw (track 9); In(2LR)elfi. bAdh' en bw/CjO (track 10); (e) HindlII-digested D N A probed with the 4.2-kb Hind111 fragment of Xob18.42: Adh"J' en (track 1 I): D f ( 2 L ) d 1 4 , Adh"f' cn/CyO (track 12); (0 DNA probed with the 9.6-kb EcoRI fragment of Xob17.36: Xhol-digested Df(2L)b83d2Ya, In(2L)b83d2Yu/SM5DNA (track 13); EcoRI-digested Df(2L)b83d2Ya, In(2L)b83d2Ya/ SM5 DNA (track 14); Xhol-digested In(2L)b83622/SM5 DNA (track 15); EcoRI-digested In(2L)b83622/SM5 DNA (track 16).

possible that the breakpoint is close to the EcoRI site at -207.6 and thefusion fragments with SalI are both the same size. This breakpoint is confirmed by in situ hybridization to polytene chromosomes, using Xob18.42 as a probe (Figure 1 Id). T h e translocation T(2;3)shv19 is broken in the 8.2-kb SalI fragment of Xob18.42 (Figure sa). When DNA from T(2;3)shu19/ C y 0 digested with EcoRI is probed with this fragment a single fusion fragment of 7.9 kb is found (Figure 10, lane 5). With SalI-digested DNA, fusion fragments of 3.7 kb and 14.0 kb are seen (Figure 10, lane 6). Therefore this translocation is broken close to one of the EcoRI sites between the limits -199.9 and -208.1. This is also confirmed by in situ hybridization (Figure 1If). In(2L)el9 is also broken in the 8.2-kb SalI fragment giving fusion fragments in a SalI digest of 6.1 kb and 11.0 kb (Figure 10, lane 10). N o fusion fragments are seen in an EcoRI digest (Figure 10,lane 8). This inversion is broken between - 199.9 and -208.1. Also broken in this region are a 1(2)35Aa+pu-eldeletion, Df(ZL)A400, and a 1(2)35Au-pu- el- deletion, Df(2L)TE146(Z)GW7.DNA from Df(2L)A400/ Cy0 nB digested with XhoI and probed with the 5.0-

kb EcoRI fragment of Xob21.I 1 (Figure 2) has the parentalfragments of 8.0 kb and 5.6 kb andan additional 6.5-kb fusion fragment(Figure10, lane 12). T h e deletion is not broken in the 8.0-kb fragment (data notshown) so is therefore brokenin the adjacent 5.6-kb XhoI fragment between -238.6 and -243.5. Df(2L)TE146(Z)GW7/CyO DNA has a fusion fragment of 0.8 kb when probed with the 2.8-kb EcoRI fragment of Xob22.35 (Figure 2) and is thus broken between -247.0 and -247.8 (Figure 10, lane 14). The breakpoints of two translocations map genetically between the 1(2)35Aaand puloci, T(Y;2)GT2 and T(2;3)GT3. Both of these are y-ray-induced translocations, and neither are mutant for pu or 1(2)35Aa (GURB et al., unpublished data). DNA from T(Y;2)GT2/Df(2L)b84~2 digested with EcoRI and probed with the 9.0-kb EcoRI fragment of Xob20.43 (Figure 2), lacks the parental fragment of 18.0 kb and has fusion fragments of 19.5 kb and 28.0 kb (Figure 10, lane 16). Only the 19.5-kb fusion fragment is seen with Xobl9.48 (data not shown) therefore this translocation breaks between the endof Xobl9.48 (-220.9) and the EcoRI site at -226.0 (see Figure 1 I C for in

113

Analysis of the el-noc Complex o!o

13

18.0

8

j g

$8

2.7

I

1.7

I

I

2.5

8.2

S

I

I I

C y e Ln

F

5.2

I

1.9

I

.-

I

1.9

I

1

5.4

S

I

8.5

xb

I

H

M,

1

4.2

H 2.7

3.6

o! rm

f

I I

2.7

I

1.7

I

2.5

I

1.6

I

2.5

I

5.4

I

4.6

I

4.6

ob18.41

I0 ~ & 1 8 . 4 2

5.2

I

5.2

1.9

I

5.4

o!

I

I

p

9.6

I

I

9.3

5.4 20.7

9.6

I

' H'PI

12.5

I 2.9

I 2.1 I

8.0

I

8.5

I

4.4

37.0

5.2

P

I

5.4

1

0.7

5.2

E

s X

ob16.19

1y

ob1736

fbl FIGURE8,"Expanded maps of the regions covered by (a) the phage Xob17.41,Xob18.41 and Xob18.42. (b) The phage Xob16.19 and Xob17.36. The coordinates are shown above the EcoRI (E) sites, only relevant Hind111 (H), Sal1 (S), XbaI (Xb) and Xhol (X) sites are shown.

situ results). When DNA from T(2;3)GT3/Cy0 is

probed with the 5.6-kb EcoRI/HindIII fragment of Xob21.11 (Figure 2),EcoRI fusion fragments of 5.3 kb and 8.8 kb are found (Figure 10, lane 18). Thus the breakpoint of T(2;3)GT3 is between -234.4 and -240.0. The y-ray-induced translocationT(2;3)GTIOis broken in the 6.5-kb SalI fragment of Xob22.35 (Figure 2) giving SalI fusion fragments of 6.2 kb and 11.5 kb (Figure 10, lane 20) and is therefore broken between -253.5 and -260.0. This translocation is broken distal to the locus 1(2)34Fd (Figure 12). Distal to Xob22.35 a further 100 kb of DNA have been cloned in overlappingphage(Figure 2). The only aberration that has been mapped to this DNA is the translocation T(2;4)GT6.This translocation falls between the loci E(2)34Fband l(2)34Fc(Figure 1). By in situ hybridization to polytenechromosomes, the breakpoint of T(2;4)GT6 is broken within the DNA covered by thephage Xob22.35 and Xob22.29, i e . , between -241.0 and -273.8 (data not shown).

DISCUSSION

A summary of the molecular relationships of the aberrations mapped in this study to each other and to the genetic loci is given in Figure 12. The relationship of noc to 1(2)35Ba:T h e molecular limits of the noc locus are between the noc+ ospdeletion Df(2LYn52 at -52.8 to -57.4 (proximal limit) (CHIA et al. 1985a)andthe 1(2)35Ba- nocdeletion D f ( ~ ) A d h at " ~-1~ 19.5 ~ to -1 2 1.8 (distal limit) (MCGILL1985). However, the chromosome 2 breakpoint of the translocation T(2;3)Mpe is between -98.5 and -103.0 (CHIAet al. 1985a) yet this translocation is completely wild type for no-ocelli. In other words a translocation which breaks within sequences required for noc function is not itself mutant for noc. This apparent contradiction was explained as being due to a dominant suppressor of noc on the chromosome associated with T(2;3)Mpe. However,amore detailed analysis of the noc locus (MCGILL 1985; MCGILLet al. 1988) indicates that noc does not form

T. Davis, J. Trenear and M. Ashburner

114 1

2

3

4

5

20.7 16.0

20.; 12.! 1 O.! 8 .!

6 7 8

->*

9.5 H 8.5

P

5.4

(8)

(b)

FIGURES."Southo-n transfers of restriction enzyme-digested DNA. The sizes i n kb of wild-type restriction enzyme fragments are inclic-;lted;f indicates a fusion fragment. The Cy0 chromosome hm the ~ml~nlorphic IiroKI site a t -1 64.4, Sal1 site a t -163.4 and Xhol site ;It - 17 I .6. (a) Probed with the 5.2-kb BcoRI fragment of Xobf6.19: EcoRI-digested In(2LR)DTD12R/CyO D N A (track 1); HindIII-digested In(2LR)DTDf28/CyO D N A (track 2); SalI-digested ln(2LR)DTDf2R/CyO D N A (track 3): Xhol-digested ln(2LR)DTDf28/CyOD N A (track 4); (b) probed with Xobf6.19: BcoRIdigested T(Y;2)A8U/Df(2f.)b84a2,Adh' p r p k en sp D N A (track 5): Hindlll-digested r(Y;2)ARU/Df(2L)bR4a2,Adh' pr pk cn s p D N A (track 6); BcoKI-digested Df(2L)ARRI. y+ac+/CyO D N A (track 7); Hindlll-digested Df(2L)ARRf,y+ac+/CjO D N A (track 8).

a single discrete complementation group, but that it can be divided into three separateregions nocA, nocB and nocC based upon phenotypic strength of mutations. The most distal of these is nocA. Two alleles of nocA have been isolated, Zn(2LR)noc4 (at -109.2 to -1 14.7) and (due to the insertion of the large transposable element TE146(Z) into the noc region at -107.5) (CHIAet al. 1985a). T h e strong noc phenotype of ~ o c ~ ~ is. ' thought ~ ~ ~ ( to ~ )be due to a position effect of the insertion of the transposable element, as noc+ revertants can be isolated which are incomplete deletions of the transposable element, retainingfrom 3 to 10 kb of fold-back DNA at the insertion site (CHIAet al. 1985b). T h e molecular limits of nocA are between the insertion site for TE146(Z)at -107.5 (MCGILL1985) and the breakpoints of the deletions Df(2L)A446 and Df(2L)Adh""" (-1 13.5 to -1 17.5 and -1 19.5 to -121.8, respectively). T h e loci nocB and nocC are located proximal to the breakpoint of T(2;3)Mpe. These conclusions are supported by the observation that the synthetic deletion T(2;3)MpeDH16" is phenotypically weak no-ocelli when heterozygous with not deletions or noc alleles.

Alleles of nocA have a strong no-ocelli phenotype when homozygous or when heterozygous with deletions that include nocA. T h e locus 1(2)35Ba was originally defined on the basis thatthe EMS-induced mutation 1(2)35BaAR" maps within the deletion Df(2Lzfn2, which is mutant and ASHBURNER for the loci el to Adh (WOODRUFF 197913). More detailed analysis showed that Z(2)35BaAR'"maps distal tothe breakpoints of the 1( 2)35Ba+noc- deletions Df(2L)A446 and Df(2L)A178 (ASHBURNER,AARON and TSUBOTA 1982). At the molecular level 1(2)35Bais defined by the distal breakpoint of the deletion Df(2L)A446 (above) and the proximal breakpoint of the deletion Df(2L)b84a2 (-12 1.8 to -1 25.1). Therefore the areas of uncertainty of the loci nocA and 1(2)35Ba overlap to a considerable extent (Figure 13). T h e translocation T(2;3)GT8 is lethal when heterozygous with alleles of 1(2)35Ba, suggesting that this is an allele of 1(2)35Ba, but it has a strong no-ocelli phenotype when heterozygous with 1(2)35Ba+ nocdeletions, e.g., Df(2L)A178 (our unpublished data). T h e breakpoint of T(2;3)GT8is between -1 18.0 and - 1 19.5, within thearea of overlap of nocA and 1(2)35Ba (Figure 13). One possible explanation of these data is that the phenotype of T(2;3)GT8 is due to a position effect of the translocation onone orboth of the loci nocA and 1(2)35Ba. An alternative explanation, however,is that nocA and 1(2)35Ba are a single transcription unit. It would then follow that mutations that disrupt or delete the distal end of 1(2)35Ba/nocA are lethal (e.g., Df(2L)Adh"7R'3,T(2;3)GT8)but mutations that disrupt or delete the proximal end of this gene are viable and phenotypically no-ocelli (e.g., Df(2L)A178,Zn(2LR)nocA). With respect to these two groups of mutations, Df(2L)A446 is intermediate, both in respect to the position of its breakpoint (Figure 13) and by the fact that it is semilethal with 1(2)35Ba alleles or when heterozygous with deletions which include 1(2)35Ba. The model that 1(2)35Ba and nocA might be identical is supported by the presence of atranscript starting close to theEcoRI site at -1 19.5 and extending in the 3' direction to about- 1 12.0. T h e direction of transcription is distal to proximal (Buo MENG, unpublished data) (Figure 13). If this is the transcript of the 1(2)35Ba/nocA locus, then T(2;3)GT8 would be broken within the 5' region of the transcript. T h e breakpoint of the deletion Df(2L)A446 is in, or near, the 3' end of the transcript, whereas Df (2L)A178 is proximal but close to the 3' end of the transcript (Figure 13). A thorough search for transcripts in this region has been made and only the onetranscript has been found (B. MENG, W. CHIAand M. ASHBURNER, and colleagues, unpublished data). Further evidence for this model is that the locus 1(2)35Ba has not been

1

18.0

2

Analysis of the el-noc Complex 37 4

6

5

-

2 0 118 1191753119111816420

"

-

B 8.0 W f

5.2

2.7 2.5

-9

8.2

-

1.9 1.7

115

.f.

"

-if 5.6 Y

f-

-

-1 f

*

18.0

f

5.6 O,

1.9

1

(a)

FIGUREIO.--Southern transfers of restriction enzyme-digested DNA probed with X phage or purified DNA fragments covering the region -200.0 to -250.0. The sizes in kb of wild-type restriction enzyme fragments are indicated: f indicates a fusion fragment. (a) Probed with Xob18.42: EcoRI-digested b Adh"" pr cn bw DNA (track I); EcoRl-digested T(Y;2)e14, b Adh"" pr cn bw/Df(2L)b84a2, Adh' pr pk cn sp DNA, (track 2); (t)) probed w i t h the 8.5-kb Xbal fragment of Xob18.42: EcoRI-digested T(Y;2)A15, B'y+ In(2LR)PuJJ,SD'" It Pu/Df(2L)b84a2. Adh' pr pk rn sp D N A (track 3); Sail-digested T(Y;P)A15, B ' y+ In(2LR)Pu'.", SD"" it fuJ~"/Df(2L)b84a2.Adh'pr pk cn sp DNA (track 4); (c) probed w i t h the 8.2-kb Sail fragnlent of Xobl8.42: EcoRI-digested T(2;3)shuIY/Cy0 DNA (track 5); Sall-digested T(2;3)shulY/C:yC.,DNA (track 6); KcoRI-digested Adh"" cn D N A (track 7); Sall-digested AdPf3cn DNA (track 8); EcoRIdigested In(2L)elY,AdPf7cn/CyO DNA (track 9);Salldigested ln(2L)eiY, Adh"f3cn/(;yO DNA (track 10); (d) probed with the 5.0-kb EcoRl fragment of Xob2l. 11: Xhol-digested b Adh' cn bw DNA (track I I): Xhol-digested Df(2L)A400, b cn bw/CyO DNA (track 12); (e) probed with the 2.8-kb EcoRI fragment of Xob22.35: EcoRldigested a1 dp b TI:146(Z) pr 42) pwn cn/CyO DNA (track 13); EcoRI-digested Df(2L)TE146(Z)GW7, a1 dp b pr l(2)pwn cn/CyO DNA (track 14): (r) probed with the 9.0-kb BcoRI fragment ofXob20.43:EcoRIdigested z wJJh4; + DNA (track 15); EcoRI-digested z wJJR';T(Y;2)GT2/Df(2L)b84a2, Adh' pr pk cn sp D N A (track 16); (g) probed with the 5.6-kb Hind111 fragment of Xob2I. I I : EcoRI-digested z wJJf";+ DNA (track 17); EcoRIdigested z wJJJ',';T(2;3)GT3/CyO DNA (track 18): (11) probed with the 6.5-kb Sal1 fragment of Xob22.35: Sall-digested z w J J J x+; DNA (track 19); Sall-digested z d l J x : T(2;3)GTIO/CyO DNA (track 20).

separated from nocA by any inversion or translocation event, i.e., no 1(2)35Ba- nocA+ aberrations have been found. Further, all deletions that affect one but not both of thesefunctions are 1(2)35Ba+nocA-, no 1(2)35Ba- nocA+ deletions are known. According to this model the pleiotropy of "1(2)35Ba/nocA" mutations is due to the temporal or spatial specificity of this gene's expression. For example,mutations which result in the loss of gene function during larval development could lead to lethality, mutations which lead to the loss of gene function only later in development could lead to the ocellar-phenotype. The el locus and its relationship to pu: T h e el locus was originally discovered by E. M. WALLACE in 1935 (BRIDGESand BREHME1944) as a spontaneous mutation (el'). Phenotypically, elbow flies have small, bent wings due to the absence or reduced size of the posteriorcompartment. The alulae are reduced in size with a reduced number of marginal bristles, and are often fused with the wing blade. Nineteen further alleles of el have subsequently been isolated (1 1 EMSinduced, one X-ray-induced and eighty-ray-induced).

When heterozygous with a deletion of the entire el region, el alleles vary in their phenotypicseverity from very strong alleles, such as el' and T(2;3)el24, to those which are almost wild type, such as T(Y,-2)AI5 and Zn(2L)el9. These alleles also vary in their interractions with Sco, some [e.g. T(Y;2)A15],strongly enhance the (e.g., el'),have no effect Scutoid phenotype, but others TSUBOTA and on theScutoid phenotype (ASHBURNER, WOODRUFF 1982; M. ASHBURNER, J. ROOTE, G. JOHNSON,D. KIMBRELLand T. DAVIS,unpublished data). Genetically all el mutations behave as if they fall into J. a single complementation group (M. ASHBURNER, ROOTE, G. JOHNSON,D. KIMBRELL and T. DAVIS, unpublished data). At the molecular level, the el alleles that are associated with aberration breakpoints are found to map to two clusters. The distal cluster includes the aberrations T(Y;2)el4, T(Y;2)A15, In(2LR)el6,T(2;3)shvZ9 and Zn(2L)e19, all of which are dominant enhancersof the Scutoid phenotype. T h e proximal cluster includes the aberration T(2;3)el24 and the spontaneous allele el', neither of which enhance the Scutoid phenotype. T h e aberrations T(Y;2)A80and In(2LR)DTDZ28 have

116

T. Davis, J. Trenear and M. Ashburner FIGURE 1 1 .-Hybridization in situ to the polytene chromosomes of aberrations broken in the el region. (a) T(Y;2)e/4/CyO hybridized with Xo618.42. There are three sites of hybridization to region 35,one to the Cy0 chromosome (Cy) and the other two to the Y"2" element of the translocation and the 2"Y" element of the translocation. This translocation is broken approximately one-third of the way along this phage. (b) In(2LR)DTDl28/CyO hybridized with Xob16.19. There are three sites of homology to this phage; one to the Cy0 site (Cy) and the others to the DTDI28'. and the DTDl28' sites. (c) T(Y;2)GT2/ln(2LR)Gla hybridized with Xob20.43. There are threesites of hybridiwtion to region 35. The Y"2" and the YY" elements are labeled to equal extents showing that the aberration is broken in the middle of this phage. (d) T(Y;2)Al5/cn bw hybridized with Xob18.42. Three sites are seen; the site o n the cn bw chromosome and the sites due to the Y"2" and the 2"Y" sites. This translocation is broken approximately onethird of the way along this phage. (e) T(Y;Z)ASO/CyO hybridized with Xob16.19. Three sites of hybridi~ation are seen; the CyO site (Cy), the Y"2" and the YY"sites. (9 T(2;3)shv19/CyO hybridized with Xob18.42. There are three sites of hybridi~ation; the Cy0 site (Cy), the Y 3 " part of the translocation and the site due to the fusion of 35h to 22F (the new order of these chromosomes is 2122F/35A-22F/97A-61; IOOF-97A/ 35A-60F).

their breakpoints between these two clusters, bothare phenotypically elbow+, and both fail to enhance the Scutoid phenotype. These data could be explained in several ways, of which the most parsimonious is that there are two elbow genes. Genetically this can be most obviously seen from the facts that: T(Y;2)A80is el+, yet synthetic deletionsextendingboth proximally (e.g.,T(Y;2)A8ODTE146(Z)GR154 and distally (e.g.,T(Y;2)e1gD A 8 0 4 from the T(Y;2)A80 breakpoint are both phenotypically elbow when heterozygous with el alleles. T h e only anomalous el mutant is T(2;3)e124, whose breakpoint maps to a region ofthe DNA (-1 26.0 and -1 3 1.4) not required forel function. Presumably this translocation is associated with a second lesion, close to that of el'. T h e model that el comprises two genes, a proximal elA and a distal elB, is supported by geneticdata fromcombinations of deletions. T h e heterozygote

Of(2L)b83d29a/Df (2L)TE36(R)GW8 is phenotypically elbow+, yet both of these deletionsare phenotypically elbow when heterozygous with el alleles (ASHBURNER et al., unpublished data). The el alleles that are associated with aberrations can be allocated to either elA or elB on the basis of phenotypic strength. The strong alleles map to theelA region and theweak alleles map to theelB region. T h e elA locus is defined by the deletions Of(2L)A26Oand Of (2L)A266at - 144.4 and thetranslocation T(Y;2)A80at -167.2 to -172.9. T h e elB locus is defined by the translocation T(Y;2)A80 (-167.2 to -1 72.9) and the translocation T(Y;2)GT2 (-220.9 to -226.0) (Figure 12). Flies which are heterozygous for alleles of elA and elB are phenotypically elbow. This could be explained by a model based on structural interactions between the products of the two genes. For example the protein products o f e l A and elB might form a heterodimer

Analysis of the el-noc Complex

-

- 2-6205 0

I

-220

-230

-240 I

I

-200

-210 I

I

- 1-9108 0 I

I

I

-170

- 1-61-051-04103 0

1

I

-120

wel24uGT8

w GT2

-GT3 GTlO

' rU hD v 1T 9D 1 2 8 L e19 t (e / 6

distal

H A 8 0

fn3 A245 TE146(2)GR18 $15'

+ - t

W 6414 H A15

+ A260 A266 + TE146(Z)GWl b TE146(Z)GW6 + e128

I

10 kb

n

t

:

I

;

. . n

I

I I L

b84a2

4

b83dZSa

4

117

:

b

*

;

el1 TE36(R)GW8

ARRl

+ e116

:

& A400

'

I

rw; t

I

m :

c

I

:

k

'

'

H "

1(2)35Ba

elB

l(2)35Aa

:

elA

PU

:

FIGURE 12,"The relationship between the genetic and molecular maps of the el region of chromosome 2L. The limits of the deletions and the breakpoints of other aberrations are shown, with the limits of uncertainty (narrow lines). The minimum limits are indicated by the bold lines (see the text).

distal

4.6

I

12.3 I

-

3

Y

5 : ,

P

:

m v !

9

13.0

I

10.5

I

I

R fn3

I :

t

b n78U

I

I I

1

t

b A446

b A178

t

1(2)35Ba

I

MP

noc4

I nod

I

FIGURE13,"The relationship between 1(2)35Ba and nocA. The limits of nocA are defined by the insertion point of TE146 (nocA-) at -107.9 and the 1(2)35Ea- nocA- deletion Df(2L)AdP7'" at - 1 19.5 to -1 21.8. The limits of 1(2)35Ba are defined by the 1(2)35Ba+ nocA- deletion Df(2L)A446 at - 1 13.5 to -1 17.5 and the 1(2)35Ba' nocA+ deletion Df(ZL)684a2 at -125.1 to -121.8. The amount of overlap of the regions of uncertainty of these loci is 8 . 3 kb. The putative 1(2)35Ea/nocA transcript is indicated by a bold line, the direction of transcription is shown by the arrow (BUO MENG, unpublished data). The narrow lines indicate the limits of uncertainty of the aberration breakpoints. Only EcoRI sites are shown.

(Figure 14a). In eZA/+ flies some of the eZB+ product would be complexed with altered elA product and be inactive. Nevertheless the amount of active heterodimer is 50% of that of a wild-type fly (Figure 14b). If a fly is also heterozygous for elB, the amountof active heterodimer would decrease to 25% of the wild-type level (Figure 14h). T h e elbow phenotype would then result from the amount of wild-type heterodimer de-

creasing below thethresholdrequiredfornormal function.This would predictthatboth Df(2L)elBelA-/+ and Of(2L)elB- elA+/elB+Df(2L)elA- genotypes would be wild type, since both would make 50% of the normal heterodimer and no inactive mutant heterodimer (Figure 14, d and e). This is indeed the case, no elB-elA-deletion has anydominant phenotype and the heterozygote Df(ZL)b83d29a/Df (2L)TE36 (R)GW8is wild type. In elAIDf(2L)elB flies the elbow phenotype resultsbecause only 50% of the elB product is made. This complexes with an equal mixture of the wild-type and mutant eZA products, so less than 50% of the wild-type complex is made and the el phenotype results (Figure 140. Thisis similar for eZB/Df (2L)elA flies (Figure 14g). This model is similar to that proposed by REGANand FULLER(1988) to explain the failure of complementation of mutations of the testis specific genecodingfor&tubulin ( 8 2 ) by point mutations of the gene haywire (hay), but not by hay""'l mutations. These workers have now shown that the products of the a-tubulin and &tubulin genes in Drosophila testis do form a heterodimer and that mutations of one genefail to complement mutations of the other in this way (HAYSet aZ. 1989). T h e pu locus was discovered in 1925 (MORGAN, BRIDGESand STURTEVANT 1925).In pupal flies the wings do not expand aftereclosion of the adult. T h e five known pu alleles allfall into a single simple complementation group. T h e pu locus maps genetically to be distal to eZA, on the basis of its inclusion within the synthetic deletionT(Y;2)A80DTE146(Z) A@ but not within the synthetic deletion T(Y;2)A8~TE2#6(Z)GR28p(ASHBURNER,TSUBOTA and WOODRUFF 1982 and unpublished data). Both of

118

T. Davis, J. Trenear and M. Ashburner e/A

gene products

I

Genotype

e /A*

elf?

el8

DEB3 %I Products

I

~~

Percentage wild-typ heterodirner

Phenotype

wt wt ( c ) e/B'

+

DX DX DX

1+

(d) Df(2L)elB elA

+

50

wt

50

wt wt

(f

1 Df(2L)elB

(4)+

+ / + elA *

Df(2L)elA / el6

+

25

el

FIGURE14.-Schematic representation of the intereactions between the el genes. The products of the genes and the mutant products are shown at the top of the figure as the different boxes, elA and elB are the wild type and elA* and elB* are the mutant products. The relative concentrations of the products is indicated by the size of the boxes. The percentage of wild-type heterodimer quoted is with respect to the amount present in (a).

I 25 el el

these deletions are el-, therefore pu must map between the loci elA and elB. The pu locus is probably not a vital locus since flies homozygously deleted for pu survive, albeit weakly, e.g., Df(2L)b83d29a/ Df(2L)A400 heterozygotes (our unpublished data). At the molecular level the pu locus is defined by the distal limitof the pu+ elA- deletion Df(2L)e114 at - 194.8 to- 199.9 and theproximal breakpoint of the deletion associated with the elB- translocation T(Y;2)e14 at -204.6 (Figure 12). This locus covers an interval of only 4.7 kb to 9.8 kb. On the basis of the allele Zn(2LR)e16, which maps proximal to pu but is an elB allele, it is possible that elB overlaps the pu locus, though the phenotype of In(2LR)e16 may be due to a position effect. The locus 1(2)35Aa:This locus is not partof the elnoc complex and is included here because this locus indicates the extreme distal end of the complex. The proximal limit of 1(2)35Aa is defined by the distal breakpoint of the 1(2)35Aa- deletion Df(2L)distal TE146(Z)GW7 at -247.0 to -247.8 and the breakpoint of the 1(2)35Aa+ deletion Df(2L)A400 at -238.6 to -243.5. T h e distal limit of this locus has not been defined but must be proximal to -260.0 on the basis that the translocation T(2;3)GTIO maps at -253.5 to -260.0 and breaks distal tothe locus 1(2)34Fd which is distal to 1(2)35Aa (Figure 12). Also mapping between -247.0 and -260.0 is the male sterile locus ms(2)34Fe (Figure 12). This work was supported by a programme grant from the MedicalResearchCouncil to M.A. We thank our colleagues DAVID GURU, PAUL LASKOand JOHN ROOTE for practical help and for many discussions.We thank BUO MENGfor allowing us to quote his DURRANT, unpublished data. Thanks go to PAMANGEL,BELINDA

JOHNSON and JILL REGAN for practical help with fly stocks. GLYNNIS

L I T E R A T U R E CITED AARON,C. S., 1979 X-ray induced mutations affecting the level of the enzyme alcohol dehydrogenase in Drosophilamelanogaster: frequency and genetic analysis of the null-enzyme mutants. Mutat. Res. 63: 127-137. ALEXANDROV, M. V., 1986 Genetic and cytogenetic mapping of the chromosome rearrangement breakpoints induced by highand low-LET radiations around the black locus of D. melanogaster. Drosophila Inform. Serv. 63: 21-22. 1982 The genetics ASHBURNER, M., C. S. AARONand S. TSUBOTA, of a small autosomoal region of Drosophila melanogaster,including the structural gene for alcohol dehydrogenase. V. Characterization of X-ray-induced Adh null mutations. Genetics 102: 421-435. ASHBURNER, M., S. TSUBOTA and R.C. WOODRUFF, 1982 The genetics of a small autosomal region of Drosophila melanogaster, including the structural gene for alcohol dehydrogenase. IV. Scutoid, an antimorphic mutation. Genetics 102: 401-420. BENDER,W.,P. SPIERERand A. SPIERER,1983 Chromosomal walking and jumping toisolate DNA from the Ace and rosy loci and the Bithorax complex in Drosophila melanogaster. J. Mol. Biol. 168: 17-34. P. PFEIFER, E. B. BENDER, W., M. AKAM,F. KARCH,P. A. BEACHY, LEWISand D. S. HOGNM, 1983 Molecular genetics of the Bithorax complex in Drosophila melanogaster. Science 221: 2329. BENYAJATI, C., A. R. PLACEand W. SOFER,1983 Formaldehyde mutagenesis in Drosophila. Molecular analysis ofADH-negative mutants. Mutat. Res. 111: 1-7. BRIDGES, C. B., and K. S. BREHME, 1944The mutants ofDrosophila melanogaster. Carnegie Inst. Wash. Publ. 552, p. 75. 1985a CHIA, W., R. KARP, S. MCGILLand M. ASHBURNER, Molecular analysis of the Adh region of the genome of Drosophila melanogaster. J. Mol. Biol. 186 689-706. CHIA,W., S. MCGILL,R. KARP,D. GUBBand M. ASHRURNER, 1985b Spotaneous excissionof a large transposing element of Drosophila melanogaster. Nature 316 81-83.

Analysis of the el-noc Complex

G u m , D., M. SHELTON.J. ROOTE,S. MCGILLand M. ASHRURNER, 1984 The genetic analysis of a large transposing element of Drosophila melanogaster. T h e insertion of a w+ rst+ T E into the ck locus. Chromosoma 91: 54-64. GURB, D., J. ROOTE,G. HARRINGTON, S. MCGILL,B. DURRANT,M. SHELTONand M . ASHBURNER,1985 A preliminary genetic analysis of TE146, a very large transposing element of Drosophala melanogaster. Chromosoma 92: 116-123. GUBR,D . , J . ROOTE, S. MCGILL,M . SHELTONand M. ASHBURNER, 1986Interactions between white genescarried by a large tr;tnsposing element and the zeste' allele in Drosophila melanogaster. Genetics 112: 551-57.5. HAYS,T. S., R. DEURING,B. ROBERTSON, M. PRouTand M. FULLER, 1989 Intel.acting proteins identified by geneticinteractions: ;I n~issence mutation i n a-tubulin Fails to complement alleles of rhe testis specific p-tubulin gene in Drosophilamelanogaster. Mol. Cell. Uiol. 9 875-884. KRIVSHENKO.J., 1959 New mutations report. Drosophila Inform. Scrv. 33: 95-96. I.INDSI.E.Y, I). L., LA. SANDLER,B. S. BAKER,A. T . C. CARPENTER, R . E. DENELL, J. C. HALL,P. A. JACOBS,G. L. GABOR MIKLOS, A. HESSLER,M. B. K. DAVIS, R. C. GETHMANN, R. W. HARDY, MILLER,H. NOZAWA, D. M. PARRYand M.COULD-SOMERO, 1972 Segmental aneuploidy and the genetic gross structure of the Drosophila genome. Genetics 71: 157-184. MCGILL,S . , 1985 Molecular studies of the Adh region of Drosophilamelanogaster. Ph.D. thesis, University of Cambridge, England. MCGILL,S., W.CHIA, K. KARP and M . ASHBURNER,1988 T h e molecular analysis of an antimorphic mutation of Drosophila melanogaster, Scutoid. Genetics 119: 647-661.

119

MANIATIS,T., E. F. FRITSCHand J. SAMBROOK,1982 Molecular Cloning. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. MANIATIS,T . , R. C . HARDISON,E. LACY,J. LAUER, C. O'CONNELL, D. QUON,G. K. SIMand A. EFSTRATIADIS, 1978 T h e isolation of structural genes from libraries of eucaryotic DNA. Cell 15: 687-701. MORGAN,T . H., C. B. BRlDGEsand A. H. STURTEVANT, 1925 T h e genetics of Drosophila. Bibliog. Genet. 2: 232. O'DONNELL,J., H.C.MANDELL,M. KRAUSS and W. SOFER, 1977 Genetic and cytogenetic analysis of the Adh region of Drosophila melanogaster. Genetics 86: 553-566. REGAN, C. L., and M. T . FULLER,1988 Interreacting genes that affect microtubule functions: the nc2 allele of the h a p i r e locus fails to complement the testis-specific &tubulin gene of Drosophila. Genes Dev. 2: 82-92. SEGAL, M., and W. GELBART,1985 Shortvein, a new component of thedecapentaplegicgene complex in Drosophilamelanogaster. Genetics 109: 1 19-143. SMOLIK-UTLAUT, S. M. and W. M. GELBART,1987 T h e effects of chromosomal rearrangement on the zeste-white interaction in Drosophila melanogaster. Genetics 1 1 6 285-298. WOODRUFF, R. C., and M. ASHBURNER,1979a T h e genetics of a small autosomal region of Drosophilamelanogaster, including the structural gene foralcohol dehydrogenase. I. Characterization of deficiencies and mapping of Adh and visible mutations. Genetics 92: 117-132. WOODRUFF,R. C., and M. ASHBURNER,1979b The genetics of a small autosomal region of Drosophilamelanogaster, including the structural gene for alcohol dehydrogenase. 11. Lethal mutations in the region. Genetics92: 133-149. Communicating editor: A. CHOVNICK