Abnormal Bristle Development

We report the molecular characterization of the Posterior sex combs-Suppressor 2 of zeste region of Drosophila melanogaster. The distal breakpoint of the Aristapedioid inversion divides the region into two parts. We have molecularly mapped the lesions associated with several loss of function mutations in the Polycomb group gene Posterior sex combs (Psc) proximal to this breakpoint. In addition, we have found that lesions associated with several loss of function mutations in the Suppressor 2 of zeste [Su(z)2] gene lie distal to this breakpoint. Since the breakpoint does not cause a loss of function in either gene, no essential sequences are shared by these two neighboring genes. There are three dominant gain of function mutations in the region that result in abnormal bristle development. We find that all three juxtapose foreign DNA sequences upstream of the Su(z)2 gene, and that at least two of these mutations (Arp' and vg") behave genetically as gain of function mutations in Su(z)2. Northern and in situ hybridization analyses show that the mutations result in increased accumulation of the Su(z)2 mRNA, which we argue is responsible for the bristle loss phenotype. T HE 49EF region of the second chromosome of Drosophila contains a number of genes that have given rise to interesting mutations. One of these is Posterior sex combs (Psc), a Polycomb (PC) group gene that is required not only for correct spatial expression of the genes of the BX-C (JURGENS 1985), but also for normal differentiation of the embryonic central nervous system (D. SMOUSE, N. PERRIMON and C.-T. Wu, in preparation) and normal dorsal-ventral development in the embryo (our unpublished results). A dominant gain of function phenotype associated with the Psc' mutation is suppression of the zestel-white interaction (WU 1984; Wu et al. 1989). A second gene in the region is Suppressor 2 ofzeste [Su(z )2 ] (KALISCH and RASMUSSEN 1974). This gene has given rise to a number of alleles that are dominant suppressors of the zestel-white interaction (Wu 1984; Wu et al. 1989), a phenotype which has been shown to be antimorphic in the case of Su(z)2', the strongest allele (Wu 1984). Su(z)2' and Psc' fail to complement each others recessive lethality, which has led to speculation that these mutations are allelic. A variety of evidence [see for example Wu et al. (1989) and BRUNK and ADLER (1 990)] suggests that this is not the case and that the lethality of the Psc' +/+ Su(z)2' transheterozygotes is synthetic and a consequence of an interaction between the two dominant gain of function mutations. The molecular genetic analysis presented here provides compelling evidence that Psc and Su(z)2 are separate genes. Gelietics 128: 119-192 (May, 1991) In addition to the cytologically normal mutations described above, there are three dominant gain of function mutations in the region that are associated with cytological aberrations at 49EF. Two of these (vg" and vg6') are deficiencies that share distal endpoints at 49E6,Fl (LINDSLEY and GRELL 1968; LASKO and PARDUE 1988; CRIPPS and SPARROW 1989). The third mutation, Aristapedioid (Arp) , is associated with a P element mediated inversion whose distal breakpoint is also located at 49E6,Fl (ADLER 1984; BRUNK 1989; BRUNK and ADLER 1990). All three mutations result in the loss and/or reduction in size of bristles located on the dorsal thorax. Arp also results in a second phenotype that is the homeotic transformation of arista to tarsus (BRUNK and ADLER 1990). An analysis of revertants of the gain of function phenotypes indicates that Arp' is a gain of function mutation in the Su(z)2 gene (BRUNK and ADLER 1990). We have previously reported the molecular cloning of 49EF DNA via the transposon tagging strategy (BRUNK and ADLER 1990). In this paper we present evidence that the Su(z)2 gene ncodes a 6.8-kb mRNA. Our data show that vg", vg6' and Arp' all result in increased accumulation of the Su(z)2 mRNA, which we argue is the cause of the gain of function bristle loss phenotype. The phenotypes (particularly of double mutants) suggest that overexpression of the Su(z)2 gene affects the determinative cell divisions that give rise to the four cell types of the bristle sense organ. Arp' is a chromosomal inversion and is desig120 B. P. Brunk, E. C. Martin and P. N. Adler nated Zn(2R)Arp. Since revertants of Arp' often no longer carry Zn(2R)Arp, and can be mutant for either PSC or S U ( Z ) ~ (BRUNK and ADLER 1990) we name revertants either S U ( Z ) ~ ~ ' ~ . " or PsC\~P.Y depending on the properties of individual revertant x or y. MATERIALS AND METHODS Drosophila stocks and culture: Marker mutations and balancer chromosomes are described in LINDSLEY and GRELL (1968). Flies were grown as described previously (BRUNK and ADLER 1990). A list and brief description of the relevant Psc-Su(z)2 region mutations is given in Table 1. Embryo cuticle preparations were done as described previously (ADLER, CHARLTON and BRUNK 1989). Isolation of revertants of Psc'*': The hypomorphic Psc allele Psc14'' was recovered among the progeny of dysgenic parents (LASKO and PARDUE 1988). Thus, it seemed likely to be due to a P element insertion, and hence revertable via hybrid dysgenesis. Male P S C ' ~ ~ ~ / C ~ O (P) flies were crossed to p r cn vg/CyO (M) females. The dysgenic PSC'~~'/ Cy0 progeny of this cross were then mated to Psc'/CyO virgin females. The progeny of this cross were screened for flies with straight wings ( i .e . , non-CyO-containing progeny). Since PSC'~~*/PSC' is a lethal genotype, flies with straight wings are likely to result from the reversion of the Psc'~'~ mutation. Eight such revertants were recovered and analyzed as described in RESULTS. Plasmid library screens: The plasmid libraries were constructed by NICHOLAS BROWN in the pNB40 vector (BROWN and KAFATOS 1988) using RNA isolated from an isogenic second chromosome strain carrying the markers d p cn bw. We followed the library screening protocol which BROWN devised. DNA isolation and manipulation: Standard techniques were followed. The detailed methods used are described in BRUNK and ADLER (1 990). Reverse Northern analysis: One microgram of poly(A') RNA was added to 250 ng of random primers (MANIATIS, FRITSCH and SAMBROOK 1982), heated to 65" for 3 min, and quick cooled on ice. The labeling reaction was as follows in reverse transcriptase buffer (Pharmacia). BSA (0.1 pg/ ml), dATP, dGTP, dTTP (40 pm), dCTP (2 pm), RNAsin (Promega) (0.5 unit/&, [32P]dCTP (100 pCi) and 30 units of cloned MMLV reverse transcriptase (Pharmacia). The reaction was incubated at 37" for 1-3 hr then stopped by the addition of EDTA to 20 mM and NaOH to 0.2 N followed by an incubation at 65" for 30 min. Unincorporated nucleotides were removed by two sequential ammonium acetate ethanol precipitations. Recombinant phage encompassing the walk were digested with restriction endonucleases, fractionated on a 0.9% agarose gel and transferred to Nytran (SCHLEICHER and SCHUELL). Hybridizations were done as described earlier for genomic Southern analyses (BRUNK and ADLER, 1990) except that a hotter probe (generally greater than 4 X lo6 cpm/ml of hybridization buffer) was used. RNA isolation: The RNA isolation protocol is the phenol/chloroform extraction in high salt, high SDS buffer described in MANIATIS, FRITSCH and SAMBROOK (1982). This was followed by one or two LiC12 precipitations after which the sample was passed over an oligo-dT (Collaborative Research type 3) column to enrich for polyadenylated messenger RNA species. Particular care was taken to avoid crowding (so as not to induce a stress response) since we often used hsp83 mRNA (MASON, HALL and GAUSZ 1984) as a control for RNA loading. The RNA preparations used in the Northern blot experiments described in Table 3 were derived from two separate RNA isolations from two separate batches of larvae. Northern analysis: RNA was fractionated on a 1.0% agarose, 6% formaldehyde gel as described by MANIATIS, FRITSCH and SAMBROOK (1 982), the only modification being that the running buffer contained 6% rather than 3% formaldehyde. The RNA was then transferred to Nytran (SCHLEICHER and SCHUELL) using an electroblotter (HOEFFER). The transfer was done in 1 X TAE (1 0 mM Tris (pH 7.8), 5 mM sodium acetate, 0.5 mM EDTA) at 250 mA for 6 hr to overnight. The hybridization when using single stranded RNA probes was described by THOMAS (1980). The washing procedure was modified from DELEON et al. (1983). The blots were washed once in 1 X SSC, 0.1% SDS for 15 min at 65", once in 0.1 X SSC, 0.1% SDS for 1 hr at 65" then three times in 2 X SSC for 20 min each at 50". This was followed by a 15-min wash at room temperature in 2 X SSC containing 0.5 pg/ml RNAse A. A final 30-min wash was done in 0.2 X SSC, 0.1% SDS at 50". The hybridization and washing procedure for control Northern analyses using random primed DNA probes was followed as described by SCHLEICHER and SCHUELL for the Nytran membrane. The hybridization and washing conditions for control single stranded DNA probes (for RP49) were done as suggested by ZACHAR, CHOU and BINGHAM (1 987). The hybridization signals were quantitated by scanning the exposed film with a Gilford multimedia densitometer. Control experiments were done to establish the linear exposure range for the film (LATHAM 1988). We chose to use the hsp83 mRNA as a control for RNA loading because this RNA is long enough ( ~ 3 kb (HOLMGREN et al . 1979)) that it serves as a reasonably effective control for nonspecific degradation, which the large Su(z)Z RNA would be sensitive to. MASON, HALL and GAUSZ (1984) found the hsp83 mRNA to be present at roughly equal amounts at all developmental stages. In some experiments we have also used RP49 (O'CONNELL and ROSBASH 1984) to control for loading (see for example Table 3). In situ hybridization to tissue sections: Wandering third instar larvae were frozen in OCT (Miles, Inc) and 16-pm sections made on a cryostat. The sections were attached to lysine subbed slides and post fixed in 4% paraformaldehyde for 20 min. The sections were treated with proteinase K, acetylated with acetic anhydride, and hybridized with singlestranded RNA probes labelled with [35S]ATP for 1618 hr using standard conditions (ANGERER, STOLER and ANGERER 1988; INGHAM, HOWARD and ISH-HOROWITZ 1985). Autoradiography was done using Kodak NTB2 nuclear track emulsion and exposure times were for 1-2 weeks. The slides were developed and the sections stained with Giemsa and coverslips mounted. The slides were then examined under both brightfield and darkfield optics. For the quantitation of differences between different genotypes, slides were hybridized simultaneously with the same probe, in the same hybridization buffer, washed as a group, dipped in emulsion, the emulsion developed, and the slides stained in a group. For comparisons, slides were photographed on the same role of film (Technical Pan) for the same exposure time at the same illumination level under darkfield optics. The photographic paper was exposed for the same length of time at the same illumination intensity, and developed to Drosophila Psc-Su(z)P region