Secreted signalling molecules provide cells with positional information that organizes long-range pattern during the development of multicellular animals. Evidence is presented that localized expression of Decapentaplegic instructs cells about their position along the anterior-posterior axis of the Drosophila wing in two distinct ways. One mechanism is based on the local concentration of the secreted protein; the other is based on the ability of the cells to retain an instruction received at an earlier time when their progenitors were in close proximity to the signal. Both mechanisms are involved in axis formation.

The function of MADS-box genes in flower and fruit development has been uncovered at a rapid pace over the past decade. Evolutionary biologists can now analyse the expression pattern of MADS-box genes during the development of different plant species, and study the homology of body parts and the evolution of body plans. These studies have shown that floral development is conserved among divergent species, and indicate that the basic mechanism of floral patterning might have evolved in an ancient flowering plant.

Limb development in Drosophila depends on subdivision of the limb primordia into functional units called compartments. Cell interactions across compartment boundaries establish pattern-organizing centres that control growth and specify cell fates along the anteroposterior (AP) and dorsoventral (DV) axes of the limbs. AP subdivision of the disc primordia is inherited from the embryonic ectoderm. DV subdivision of the wing disc occurs during the second larval instar through localized expression of the apterous protein (Apterous) in dorsal cells. A third major subdivision of the wing disc into wing and body-wall compartments also occurs in the second instar. Here we show that specification of the wing primordium in early second instar depends on activity of the AP patterning system but not the DV system. These results define two distinct roles for the wingless gene: a primary role in specifying the wing primordium, and a subsequent role mediating the patterning activities of the DV compartment boundary.

Proper development of petals and stamens in Arabidopsis flowers requires the activities of APETALA3 (AP3) and PISTILLATA (PI), whose transcripts can be detected in the petal and stamen primordia. Localized expression of AP3 and PI requires the activities of at least three genes: APETALA1 (AP1), LEAFY (LFY), and UNUSUAL FLORAL ORGANS (UFO). It has been proposed that UFO provides spatial cues and that LFY specifies competence for AP3 and PI expression in the developing flower. To understand the epistatic relationship among AP1, LFY, and UFO in regulating AP3 and PI expression, we generated two versions of AP1 that have strong transcriptional activation potential. Genetic and molecular analyses of transgenic plants expressing these activated AP1 proteins show that the endogenous AP1 protein acts largely as a transcriptional activator in vivo and that AP1 specifies petals by regulating the spatial domains of AP3 and PI expression through UFO.

Proper development of petals and stamens in Arabidopsis flowers requires the activities of APETALA3 ( AP3 ) and PIS-TILLATA ( PI ), whose transcripts can be detected in the petal and stamen primordia. Localized expression of AP3 and PI requires the activities of at least three genes: APETALA1 ( AP1 ), LEAFY ( LFY ), and UNUSUAL FLORAL ORGANS ( UFO ). It has been proposed that UFO provides spatial cues and that LFY specifies competence for AP3 and PI expression in the developing flower. To understand the epistatic relationship among AP1 , LFY , and UFO in regulating AP3 and PI expression, we generated two versions of AP1 that have strong transcriptional activation potential. Genetic and molecular analyses of transgenic plants expressing these activated AP1 proteins show that the endogenous AP1 protein acts largely as a transcriptional activator in vivo and that AP1 specifies petals by regulating the spatial domains of AP3 and PI expression through UFO .

No abstract

The nubbin gene is required for normal growth and patterning of the wing in Drosophila. We report here that nubbin encodes a member of the POU family of transcription factors. Regulatory mutants which selectively remove nubbin expression from wing imaginal discs lead to loss of wing structures. Although nubbin is expressed throughout the wing primordium, analysis of genetic mosaics suggests a localized requirement for nubbin activity in the wing hinge. These observations suggest the existence of a novel proximal-distal growth control center in the wing hinge, which is required in addition to the well characterized anterior-posterior and dorsal-ventral compartment boundary organizing centers.

FIG. 3 a, An autoradiogram of the initial products of digestion showing that T5 5' -exonuclease has endonucleolytic activity. Lane 1, 5' -end 32 P-labelled oligonucleotide markers of 34 and 16 nucleotides. Unlabelled template and adjacent strands were present at slight molar excess over the 34-mer flap strand (0.2 pmol). The sequence for the flap structure used was as in ref. 4. The substrates were incubated for 30min at 37°C in 12µ1 of 25mM glycine/ KOH, pH 9.3, 5 mM MgCl2 and 1 mM DTT with different concentrations ofT5 exonuclease. Lane 2, 0.003 pmol enzyme; lane 3, 0.03 pmol enzyme; lane 4, 3.0 pmol enzyme; lane 5, 0.3 pmol enzyme. Reaction products were electrophoresed on a 15% polyacrylamide gel. The initial major products of digestion are 19 and 21 nucleotides in length, consistent with cleavage at the bifurcation. b, A conceptual model of how a flap structure could bind to the T5 5'-exonuclease (produced with RIBBONS 26 ). Based upon previous experimental results (refs 3, 7 and our own unpublished observations) we have made a model of the single-stranded flap DNA (ssDNA, blue) threaded through the helical arch. The placement of the a 34---.DNA was chosen with the electrostatic potential surface contour as a guide. The flap DNA structure was docked on the protein, minimizing clashes between the two molecules. The model was not energy minimized.site occupied by zinc was observed. Thus the distance between the twp metal-binding sites we observe is much greater than the usual 4A observed in nucleases with a putative two-metal-ion mechanism, and it is likely that the details of the 5' -exonuclease mechanism will be different. 1235-1246 (1994 4127-4132 (1991). 9. Kim, Y. et al. Nature 376, 612-616 (1995). 10. Sayers, J. R. J. theor. Biol. 170, 415-421 (1994 2071-2075 (1994). 12. Beese, L. S. & Steitz, T. A. EM BO J. 10, 25-33 (1991). 13. Lima, C. D .. Wang, J. C. & Mondragon, A. Nature 367, 138-146 (1994). 14. Berger, J. M .. Gamblin, S. J., Harrison, S. C. & Wang, J. C. Nature 379, 225-233 (1996). 15. Wigley, D. B., Da-Aes, G. J .. Dodson, E. J., Maxwell, A. & Dodson, G. Nature 351, 624-629 (1991 LETTERS TO NATURE 2345The two metal ions are shown as silver spheres. The precise position of the double-stranded DNA docked to the enzyme needs to be determined experimentally.A more complete understanding of the mechanism of the enzyme will clearly require co-crystallization with DNA. Our structure will assist in the devising of the site-directed mutagenesis experiments required to elucidate the mode of action for this member of a biochemically important class of enzymes. D16. Robins, P. , Pappin, D.