Copyright 8 1988 by the Genetics Society for America Molecular Analysis of the Neurogenic Locus mastermind of Drosophila melanogaster Barry Yedvobnick, David Smoller, Pamela Young and Diane Mills Department of Biology, Emoq University, Atlanta, Georgia 30322 Manuscript received August 25, 1987 Revised copy accepted November 21, 1987 ABSTRACT The neurogenic loci comprise a small group of genes which are required for proper division between the neural and epidermal pathways of differentiation within the neuroectoderm. Loss of neurogenic gene function results in the misrouting of prospective epidermal cells into neuroblasts. A molecular analysis of the neurogenic locus mastermind (mam)has been initiated through transposon tagging with P elements. Employing the Harwich strain as the source of P in a hybrid dysgenesis screen, 6000 chromosomes were tested for the production of lethal mam alleles and eight mutations were isolated. The mam region is the site of residence of a P element in Harwich which forms the focus of a chromosome breakage hotspot. Hybrid dysgenic induced mum alleles elicit cuticular and neural abnormalities typical of the neurogenic phenotype, and in five of the eight cases the mutants appear to retain a P element in the cytogenetic region (50CD) of mum. Utilizing P element sequence as probe, mam region genomic DNA was cloned and used to initiate a chromosome walk extending over 120 kb. The physical breakpoints associated with the hybrid dysgenic alleles fall within a 60- kb genomic segment, predicting this as the minimal size of the mam locus barring position effects. The locus contains a high density of repeated elements of two classes; Opa (CAX), and (dC-dA), * (dG-dT),. A preliminary study of the transcriptional activity of the mum region is presented. HE ventral and cephalic portions of the central (E[spQ),mmtemnind (mum),big brain (bib)and neuralized T nervous system of Drosophila derive from a (neu). A maternal classof neurogenic mutations, restricted section of the ectoderm termed the neu- including almondex (am)(LEHMAN et al. 1983), pecanex rogenic region (POULSON1950). The ventral region (pcx) and several female-sterile mutations (PERRIMON consists of cells situated lateral to the mesoderm at et al. 1986), have also been isolated. An intriguing the blastoderm stage, and flanking the ventral mid- characteristic of the neurogenic gene classis the line during gastrulation. Although the neurogenic interactions observed between certain genes (WEL- ectoderm appears homogeneous, it is composed of SHONS 1956; CAMPOS-ORTEGAet al. 1984; VASSIN, cell populations destined to form two major tissues. VIELMETTERand CAMPOS-ORTEGA1985; M. A. T. Approximately 25% of the cells will form neuroblasts, MUSKAVITCH, personal communication). For in- the precursors of the central nervous system, whereas stance, E(sp1) enhances the phenotype of a Notch the remaining 75% will differentiate into epidermal locus allele(spl), while chromosome breaks associated structures (HARTENSTEINand CAMPOS-ORTEGA1984). with reversion of the enhancer phenotype produce The choice between these two developmental paths a neurogenic lethal allele at the E(sp1) locus (termed can be disrupted through mutation. Members of the E[spQR).Further, E(~p1)~alleles do not complement neurogenic class of mutations display an embryonic Delta or Notch alleles in double heterozygous combi- lethal phenotype associatedwith a misrouting of nations, suchas E(s@ + I + Dl. In addition, epidermal precursors into neuroblasts (POULSON embryos hyperploid for certain wild type neurogenic 1950; CAMPOS-ORTEGA1985). Thus, the neurogenic loci can reduce the neurogenic phenotype of specific loci represent agroup of genes which effect a specific mutant embryos. For instance, homozygous bib, mum, developmental decision. or neu embryos can bepartially rescued by the Saturation mutagenesis screens for zygotic lethal introduction of extra wild-type Notch loci. These mutations detected six genes that are mutable to a observations suggest that particular neurogenic loci neurogenic state (LEHMANet al. 1983; JURGENS et al. share a common function and possibly comprise a 1984; NUSSLEIN-VOLHARD, WIESCHAUSand KLUDING developmental pathway. 1984; WIESCHAUS,NUSSLEIN-VOHARD and JURGENS Inan effort to determinethe function of the 1984), including Notch (N),Delta (Dl),Enhancer of split neurogenic loci, these genes have become the targets The sequence data presented in this article have been submittedto the of cloning in several laboratories. At this time, only EMBUGenBank Data Libraries under the accession numberY00611. the Notch locus has been well characterized at the Genetics 118: 483-497 (March, 1988). 484 B. Yedvobnick et al. molecular level. Notch spans approximately 38 kb of MATERIALS AND METHODS genomic DNA and encodes a poly(A+) RNA of 10.5 kb (ARTAVANIS-TSAKONAS,MUSKAVITCH and YEDVOB- Strains: Harwich and seF8, P and M strains, respectively, were supplied by M. KIDWELL.A multiply marked balancer NICK 1983; KIDD, LOCKETTand YOUNG1983; GRIM- M strain was supplied by V. FINNERTY.Males from this WADE et al. 1985). cDNAs spanning the entire 10.5 stock were outcrossed to P mutants to reduce the back- kb transcript have been synthesized and sequenced ground of P elements in the stockused for molecular (YEDVOBNICKet al. 1985; WHARTONet al. 1985b; KIDD, cloning. Mam'J'13 and bib''" were provided by E. WIES- KELLYand YOUNG 1986). Conceptual translation of CHAUS. Mam"J'4 was provided by C. NUSSLEIN-VOLHARD. Zn(2R)N2GISM5 was obtained from Cal Tech. Canton S the cDNA sequence demonstrated thatNotch appears was obtained from Bowling Green. All crosses were per- toencode a membrane-spanningpolypeptide with formed at 25" on standard corn meal and molasses medium. epidermal growth factor-like repeats in the extracel- In situ hybridization and cytology: Cloned DNA was lular domain. The putative structure of the Notch biotinylated through nick translation and hybridized to protein immediately implicated cell: cell interactions polytene chromosome preparations as described by LAN- GER-SAFER,LEVINE and WARD(1982). The hybridized probe as a component in the process of neurogenic ecto- was visualized employing a horseradish peroxidase detec- derm differentiation (WHARTONet al. 1985b; KIDD, tionsystem (Enzo Biochemicals)against Giemsa-stained KELLYand YOUNG1986). Subsequent DNA sequence chromosomes. The preparations were photographed using studies of a Delta locuscDNA suggested a similar Kodachrome film (ASA 25 or 64) and an NCB 10 filter. Cuticle preparation:Cuticles from mature embryos were function for theproduct of this neurogenic locus prepared according to the methods of VANDER MEER(1977) (VASSINet al. 1987). This hypothesis is made more and photographed using Kodachrome film (ASA 25 or 64) attractive by the observations of DOE and GOODMAN and an NCB 10 filter. (1985a,b). Employing laser ablation to eliminate in- Staining of embryonic central nervous systems: Direct dividual neurogenic ectodermal cells in grasshopper immunofluorescence of embryo whole mounts was per- formed using a protocol supplied by MARCMUSKAVITCH embryos, itwas demonstrated that neuroblasts appear (Indiana University). It is a variation of a method of T. toinhibit adjacent cells from forming neuroblasts, KARRand B. M. ALBERTS(UCSF), developed by N. PATEL thereby implicating cell contact as a component in and C. GOODMAN(Stanford University).Adults wereal- the decisionmechanism. Similar conclusions have lowed to lay eggs on heavily yeasted bottles for 8 hr at 18", beenreached by TECHNAUand CAMPOS-ORTEGA and the embryos were aged12-16 hr at 18". Aged embryos were collected on a nylon screen, washed and transferred (1986, 1987) whoreported that transplanted ecto- to silanizedglass culture tubes (12 mm X 75 mm). The dermal cells of early Drosophila gastrulaexhibit a embryos were dechorionated in 50% bleach for 5 min, and degree of nonautonomous differentiation, as well as rinsed with 0.02% Triton X-100. After aspiration of the by Hoppe and Greenspan (personal communication) rinse buffer, the embryos were fixed by suspension in 0.5 through the analysis of Notch mosaic embryos. ml4% w/v paraformaldehyde to which were added 1.5 ml heptane. The suspension was vortexed repeatedly for 10- The observationscited above have established a 20 min, and the lower phase was removed by aspiration. provisional model for how ectodermal cells within The vitelline membranes were removed through addition the neurogenic region enter the neuralor epidermal of 0.5 ml 100% methanol for 1-2 min. The solution was path of differentiation. However, the model does not aspirated and the embryos were rinsed 2 times in 1.5 ml address the role of maternal vs. zygotic contributions 100% methanol, and 3 times in 0.5 ml phosphate-buffered saline (PBS: 130 mM sodium chloride, 7 mM dibasic sodium of these genes, nor does it position the neurogenic phosphate, 3m~monobasic sodium phosphate). The em- loci within the larger framework of the regulatory bryos were rinsed in 200 p1 of PNS (PBS + 1% v/v normal hierarchy that unfolds during earlyDrosophila devel- goat serum, Worthington Cappel). This rinse was removed opment. Thus, the establishment of a cell-cell inter- and an additional 200 pl PNS was added. One microliter action is an event that occurs within a continuum. of a preparation of fluoresceinated anti-horseradish per- oxidase (rabbit sera, Worthington-Cappel) was added to This interaction must be preceded and followed by the suspension, and the tubes were sealed and