The kinase Sgg modulates temporal development of macrochaetes in Drosophila by phosphorylation of Scute and Pannier

Abstract

Evolution of novel structures is often made possible by changes in the timing or spatial expression of genes regulating development. Macrochaetes, large sensory bristles arranged into species-specific stereotypical patterns, are an evolutionary novelty of cyclorraphous flies and are associated with changes in both the temporal and spatial expression of the proneural genes achaete (ac) and scute (sc). Changes in spatial expression are associated with the evolution of cis-regulatory sequences, but it is not known how temporal regulation is achieved. One factor required for ac-sc expression, the expression of which coincides temporally with that of ac-sc in the notum, is Wingless (Wg; also known as Wnt). Wingless downregulates the activity of the serine/threonine kinase Shaggy (Sgg; also known as GSK-3). We demonstrate that Scute is phosphorylated by Sgg on a serine residue and that mutation of this residue results in a form of Sc with heightened proneural activity that can rescue the loss of bristles characteristic of wg mutants. We suggest that the phosphorylated form of Sc has reduced transcriptional activity such that sc is unable to autoregulate, an essential function for the segregation of bristle precursors. Sgg also phosphorylates Pannier, a transcriptional activator of ac-sc, the activity of which is similarly dampened when in the phosphorylated state. Furthermore, we show that Wg signalling does not act directly via a cis-regulatory element of the ac-sc genes. We suggest that temporal control of ac-sc activity in cyclorraphous flies is likely to be regulated by permissive factors and might therefore not be encoded at the level of ac-sc gene sequences.

Fig. 1.

Bristles are shorter in wingless loss-of-function and shaggy/GSK-3 gain-of-function mutants and formation of bristle precursors is delayed. (A-E) The thoraces of wild type (A), c765-Gal4 >UAS-Sgg^WT (B), c765-Gal4>UAS-Sgg^act (C) and wg^Sp/wg^CX4 (D) are shown. Arrows indicate the aDC bristle. The lengths of the DC bristles are given in E. Error bars represent mean±s.e. (F,G) Wing discs from white prepupae of wg^Sp/wg^CX4 (G) and heterozygous wg/+ control larvae (F) are stained with an anti-Hindsight antibody to reveal bristle precursors. Arrows point to the DC precursors, which are visible in the control disc but have not yet formed in the mutant disc. Arrowheads point to pDC precursors present in both wild-type and mutant discs.

Fig. 2.

Expression of scute and a DCE reporter gene in sgg mutants and after mutation of putative dTCF binding sites. (A-A″) Expression of the DCE-lacZ reporter (red) is visualized in a third-larval-instar wing disc bearing a clone of cells mutant for sgg/GSK-3^D127 (outlined). Staining for Hindsight (blue) indicates the dorsocentral bristle precursors. An additional domain of DCE-lacZ reporter expression is present in the sgg/GSK-3^D127 clone and includes an ectopic precursor. (B) Co-expression of the DCE-lacZ (red) and the DCE^T123-GFP (green) reporters, in which the three putative dTCF binding sites are mutated, is shown in a third-larval-instar wing disc. The two coincide at the position of the dorsocentral proneural cluster. (C,C′) Ectopic expression of scute (red) in a clone of cells mutant for sgg/GSK-3^M11 (outlined) is shown.

Fig. 3.

Phosphorylation of Pannier by Sgg and bristle phenotypes observed after expression of Pannier bearing mutated phosphorylation sites. (A) The amino acid code of the mouse microtubule associated protein (MAP1B) peptide sequence used to raise the BUGS phospho-antibody (Trivedi et al., 2005) is compared with two potential phosphorylation sites in Pannier (Pnr). Asterisk indicates conserved residues. The central serine residue (S, bold) is the site for phosphorylation and is conserved in Pnrβ at amino acids 101 and 391. Red indicates bases altered in the pnr mutant construct. (B) Immunoprecipitated pannier (pnr) and scute (sc) proteins were used for western blot analysis with the Pnr and BUGS antibodies. A 60 kDa band is seen with both the phospho-antibody BUGS and the Pnr antibody in duplicate filters. No signal was seen with immunoprecipitated Sc as a control. (C) Western blotting showed that incubation of His-Pnr with Sgg generates the BUGS phospho-antibody epitope. The phosphorylation signal in immunodepleted Sgg supernatant (Sgg-) was largely reduced. Treatment with the GSK3β inhibitor LiCl (30 mM) during the kinase assay reaction prevents phosphorylation of Pnr. Loading controls are shown in the lower panel using an antibody against His-tag. (D) The number of DC bristles present after overexpression of wild-type Pnr (c765-Gal4>UAS-Pnr^WT) or a mutated form of Pnr (see A) (c765-Gal4>UAS-Pnr^MUT), is shown together with images of representative flies. (E) The number of DC bristles present after overexpression of wild-type or mutant Pnr in a loss-of-function pnr mutant devoid of DC bristles (pnr^VX6/pnr^MD237>UAS-Pnr^WT and pnr^VX6/pnr^MD237>UAS-Pnr^MUT) is shown, together with images of representative flies. Pnr^MUT rescues more bristles than Pnr^WT. (F) The number of DC bristles present after overexpression of wild-type or mutant Pnr in a hypomorphic wg mutant (see Fig. 1D), is shown, together with images of representative flies (wg^Sp/wg^CX4; MD237-Gal4>UAS-Pnr^WT and wg^Sp/wg^CX4; MD237-Gal4>UAS-Pnr^MUT). Pnr^MUT rescues more bristles than Pnr^WT. Error bars represent the mean±s.e.m. from three independent experiments.

Fig. 4.

Phosphorylation of Scute by Sgg and bristle phenotypes observed after expression of Scute-bearing mutated phosphorylation sites. (A) Sequence indicating serine residues (S) at potential sites (amino acids 85 and 268, red) for phosphorylation of Scute (Sc) by Sgg. bHLH, basic helix-loop-helix domain. (B) The in vitro protein kinase assay with radiolabelled [λ-P³³]ATP is shown. Recombinant Scute was incubated with Sgg kinase in a kinase assay buffer (see Materials and methods). A positive band at about 75 kDa indicated kinase activity on the SDS-PAGE gel. No kinase activity was detected when recombinant Sc alone is used. Treatment with the GSK3β inhibitor LiCl during the kinase assay reaction prevents phosphorylation of Sc. A duplicated gel was stained with Coomassie Blue as a loading control (lower panel). (C,D) The number of DC bristles present after overexpression of wild-type Sc (DC-Gal4>UAS-Sc^WT) or a mutant form of Sc (the central serine residue of phosphorylation site 268, as well as a possible priming phosphate site located at amino acid 272, were mutated into alanines (shown in blue in A), was examined in wild-type flies (C) (DC-Gal4>UAS-Sc^WT and DC-Gal4>UAS-Sc^MUT) and in a hypomorphic wg mutant (D) (Fig. 1D) (wg^Sp/wg^CX4; DC-Gal4>UAS-Sc^WT and wg^Sp/wg^CX4; DC-Gal4>UAS-Sc^MUT). Sc^MUT induces formation of more ectopic bristles than Sc^WT. Error bars represent mean±s.e.

Fig. 5.

The gene regulatory network controlling expression of achaete-scute in the notum of D. melanogaster. Diagram outlining the interactions between genes regulating scute expression. Arrows in black indicate a transcriptional response of target genes; it is not known whether activation of wg by Pannier is a result of direct transcriptional regulation. Pannier and the Iro-C gene products activate sc expression through numerous cis-regulatory sequences. Segregation of sensory organ precursors requires an autoregulatory element, the SOPE. Scute itself binds the SOPE as well as other factors including Rel/NFκ-B. Rel/NFκ-B also indirectly affects stability of sc transcripts. Red lines indicate three phosphorylation targets of Sgg: Mad, Pnr and Sc. These proteins are less active in the phosphorylated state and all affect sc expression directly or indirectly. Mad/Dpp signalling regulates expression of pnr, ush and the Iro-C genes, Pnr and Iro-C products activate sc and Sc positively autoregulates itself via the SOPE. Wingless signalling is known to inactivate Sgg (green). Hyperactive nonphosphorylated forms of Mad, Pnr and Sc accumulate and result in increased sc activity and bristle precursor development.