Drosophila melanogaster Germ Cell Migration

We describe a novel genetic locus, wunen ( w u n ) , required for guidance of germ cell migration in early Drosophila development. Loss of wun function does not abolish movement but disrupts the orientation of the motion causing the germ cells to disperse even though their normal target, the somatic gonad, is well formed. We demonstrate that the product of this gene enables a signal to pass from the soma to the germ line and propose that the function of this signal is to selectively stabilize certain cytoplasmic extensions resulting in oriented movement. To characterize this guidance factor, we have mapped wun to within 100 kb of cloned DNA. G ERM cell migration and gonad formation in Drosophila (RABINOWITZ 1941; SONNENBLJCK 1941; COUNCE 1963) have received renewed attention using reagents that identify the germ cells (HAY et al. 1988; LASKO and ASHBURNER 1988) and the somatic gonad (BROOKMAN et al. 1992) and techniques to examine the migration both in uivo and in culture (JAGLARZ and HOWARD 1994, 1995). Pole cells, the precursors of the germ line, form at he posterior pole outside the somatic tissue at stage 4 of CAMPOS-ORTEGA and HARTENSTEIN (1985) and move into the invaginating gut primordium as gastrulation begins. The germ cells then enter the developing body by passing across the epithelium of the posterior midgut primordium (PMGP) at stage 10. After this they enter the mesoderm and associate with the primordium of the somatic gonad before the beginning of germ band retraction at stage 13. Genes required for this migration will affect at least three interrelated processes: (1) determination of the germ cells and the somatic tissues through which they migrate, (2) generation of motility itself, and (3) the orientation of this activity. To date all the described mutations in the first class are thought to act in the soma. These interfere with transepithelial migration (JAGLARZ and HOWARD 1994; WARRIOR 1994) and gonad condensation (CUMBERLEDGE et al. 1992; WARRIOR 1994; BOYLE and DINARDO 1995; GRIEC and AKAM 1995). Mutations in the second class of genes will result Cowesponding author; Ken Howard, Roche Institute of Molecular Biology, Roche Research Center, 340 Kingsland St., Nutley, NJ 07110. E-mail: [email protected] ' Present addre.ss:The Jackson Laboratory, 600 Main St., Bar Harbor, ME 04609. '' Presenl address: Department of' Virus and Cell Biology, Merck Re'* P r e s e n t uddress: Department of Biological Sciences, University of search Laboratories, West Point, PA 19486. Notre Dame, Notre Dame, IN 46.556. Genetics 143 1231-1241 (,Julv, 1996) in immotile cells, while the third class of mutations will result in cells that, though motile, will be defective in guidance. In an attempt to locate such genes, we screened a set of deficiency chromosomes from the Bloomington Stock Center by staining collections of embryos from these stocks with anti-vasa antibody and looking for both dominant and recessive effects of the deficiency on migration (N. ZHANG and K. HOWARD, unpublished data). The deficiency chromosomes Df(2R)44CE, Df(2R)cn9, and Df(2R)GR15'.LT' [an inversion chromosome that throughout this manuscript we will describe as In(2LR)GRIS1~L~] all showed a characteristic failure in germ cell guidance and identify wunen (wun), a gene belonging to the guidance class. The available data on this material suggested that wun lay in a small region in 44D on the polytene map. Here we show that in each of these chromosomes lesions in wun lie outside of the described deletions and that wun is located at 45D. We use these lesions, a deficiency and two cytologically invisible lesions, to describe the wun loss of function migration phenotype and use cell transplantation to show that wun acts in the somatic tissue. We also show that both cytologically invisible alleles display a temperature-sensitive semi-lethal phenotype. Finally, we use the genetic material described here and cloned DNA from the genome projects described in FlyBase (ASHBURNER and DRBDAL.E 1994) to map wun to within 100 kb of DNA. MATERIALS AND METHODS Fly stocks: Unless otherwise noted all stocks used in this study are described in LINDSLEY and ZIMM (1992). Flies were reared at 25" on a standard medium of yeast extract, cornmeal, and sugar seeded with live bakers' yeast. X-ray screen: Fourto five-day-old males homozygous for j d 3 , a viable P[lac, w'] insert at ptc (J. HOOPER, personal 1232 N. Zhang et al. communication), were irradiated in a Torex 120D machine running at 1.2 kV 4 mA delivering an unfiltered dose of 45 krad. About 50 irradiated males were crossed en masse with 200 ru;Sp/CyOvirgin females. These flies were raised in bottles and transferred four times at 2day intervals. wFI flies were crossed singly to zu;Sp/CyO virgins to establish stocks. These wstocks were tested for lethality over Df(2R)44CE, and those that were lethal were examined cytologically. Complementation tests: Crosses for complementation tests were set up using 10 2to May-old virgin females and 10 males. The cultures were maintained at a constant temperature and transferred to fresh vials every 3 days. All the emerging progeny were scored for Cy (crosses using CyO) or Roi (crosses using SM6, Roi eue-lazz). In all cases, all the progeny from individual vials were examined and at least 100 flies scored in total. Noncomplementation was scored if there were no transheterozygous progeny in the sample. Normally the crosses were maintained at 25". Other temperatures were used, as indicated in the text, with transfer times at 27" reduced to 2 days and at 18" extended to 4 days. Histology: Embryos were collected on nylon mesh using a 12-place Millipore Filtration manifold, dechorionted in Clorox, fixed in phase partition (ZALOKAR and ERK 1977) essentially as described by MITCHISON and SEDAT (1983). Embryos were then incubated in 1:10,000 dilution of an anti-vasa antiserum raised in chicken, and the signal visualized using secondary reagents from Jackson Immuno Research. In some cases mutant embryos were identified using the lacZ balancer chromosome SM6, Roi eve-lazZ. To visualize the somatic gonad, stained embryos were further processed using digoxigenin-labeled in situ probes to transcripts of the 412 transposon (BROOKMAN et al. 1992). For the developing gut we used either the enhancer trap line 4850 (HOWARD et al. 1993) or forkhad (WEIGEL et al. 1989) nucleic acid probes prepared by PCR. Mosaic analysis: Pole cells were transplanted from a faflac% homozygous stock where the germ cells are labeled by a transgenic lacZ fusion protein construct (FISCHER-VIZE et al. 1992) into embryos from the Df(2R) wun ';'./SM5,Roi eve lacZ stock using standard techniques. The embryos were maintained at 18" and fixed 24-30 hr later using 4% paraformaldehyde; the vitelline membranes were removed using a combination of agitation in methanol and manual manipulation before staining with anti beta gal antiserum to reveal the transplanted germ cells and the genotype of the recipient. If more than five transplanted germ cells had entered the gonad, development was scored as normal. If fewer than five had entered the gonad and at least that many cells could be identified, the specimen was scored disperse. If the morphology of the embryo was too disturbed to reliably identify both gonads or fewer than five germ cells could be identified, the sample scored N.D. Molecular biology: Standard molecular techniques were used for clone isolation and mapping. We used the A fix genomic library from Stratagene and grew YAC DNA according to CAI et al. (1994) and P clone DNA according to the protocols of Genome Systems. Plasmid rescue was performed by homogenizing 10 adult flies in a microfuge tube in 100 pl of 0.3 M TrisHCI pH 8.8, 0.1 M EDTA, 1.2.5% SDS, 5% sucrose, 0.5% DEPC. The homogenate was incubated at 60" for 30 min, cooled to room temperature, 15 pl of 8 M KOAc added, and the mixture incubated on ice for 30 min. After centrifugation for 30 min at 13 krpm at 4", 80 p1 of the supernatant was mixed with 600 pl of Qlagen buffer PB and purified on Qlagen PCR spin column. The eluate was digested with appropriate restriction enzymes, and the reaction was terminated by addition of 600 pl Qlagen buffer PB followed by further purification on a Qlagen PCR spin column. OneFIGURE 1.-The umn phenotype revealed by deficiency chromosomes. (A) Germ cells in a normal gonad (arrow) in a stage14 wild-type embryo. (B) Dispersed germ cells in a homozygous Df(2R)44CE embryo at a comparable stage showing the ptc segment polarity phenotype and dispersed germ cells. (C) A homozygous In(ZR)CRZ5'.I,Y embryo showing dispersed germ cells but no segment polarity phenotype. fifth of the digested DNA was ligated in a 10O-pl reaction incubated on melting ice overnight before another Qiagen PCR spin column purification and transformation of Nova Blue competent cells. DNA sequencing was performed by the Roche Institute Core Sequence Group using a P-element primer and an IBI sequencing system. Sequence analysis was performed using Kodak Mac Vector and Assembly Lign Software. PCR was performed on a MJ Research PTC 100 machine, using standard .3-STEP [94" for 1 min; 30 X (92" for 40 sec, 60" for 40 sec, 7.5" for 90 sec); 5 min at 75"), Perkin Elmer Taq polymerase and Boehringer PCR buffer.