Introduction
Mammalian GTP-bound active Ras binds directly to the RBDs of class I PI(3)Ks (p110
, 
, 
and 
)5, 7, 8, 10. These interactions, similarly to the interactions between PI(3)Ks, their adaptors and growth factor receptors, increase PI(3)K signalling through allosteric mechanisms and by bringing PI(3)Ks closer to their substrates at the plasma membrane1, 3, 7. A critical output of PI(3)K activation is phosphorylation and activation of the serine/threonine kinase Akt. Indeed phosphorylation of Akt on its carboxyl-terminal hydrophobic motif by the Tor–Rictor complex depends on PI(3)K activity and is a widely used readout of PI(3)K activation in vivo (see Supplementary Information, Fig. S1)16, 17. Oncogenic constitutively active Ras also interacts with PI(3)K, which is likely to have important consequences for tumorigenesis. Until now, the interaction between mammalian PI(3)K and Ras has been studied using purified or overexpressed proteins in vitro and in cultured cells. Here, we use a knock-in mutagenic approach in Drosophila to investigate the functional importance of the interaction between endogenous PI(3)K and Ras in vivo.
Drosophila possess one class I PI(3)K, Dp110, and one SH2 domain-containing adaptor subunit, p60 (Refs 14, 18). Dp110 is activated by insulin signalling and regulates growth13. Consistent with p60 being required for Dp110 activation, Dp110 and p60 have the same mutant phenotypes: null mutations are larval lethal and null-mutant clones grow slowly and have reduced cell size and cell number19. Dp110 contains a conserved RBD, suggesting that Ras1 (the Drosophila homologue of mammalian H-, N- and K-Ras) or Ras2 (the Drosophila homologue of the mammalian R-Ras family) contributes to Dp110 activation (see reference 20 for a dendrogram of Drosophila and human Ras superfamilies). Expression of constitutively active Ras1V12 in wing imaginal discs increases growth, cell size and PI(3)K signalling, as monitored by immunodetection of phospho-DAkt or the membrane localisation of a 3-phosphoinositide-binding pleckstrin homology (PH) domain-GFP fusion protein (tGPH) (see Supplementary Information, Fig. S1a, b)21, 22, 23. Although wing clones with reduced Ras1 activity are small and contain small cells22, 24, PI(3)K signalling is not detectably reduced in these clones (see Supplementary Information, Fig. S1c, d)23. ras2 mutants have not been isolated, therefore their effect on PI(3)K signalling has not been analysed; however, activated Ras2 expression has no detectable effect on phospho-DAkt levels (data not shown) and no clear growth phenotype. Instead, it affects ommatidial planar polarity and cell-fate specification25. Taken together, these data suggest that Ras2 does not activate Dp110, but that Ras1 might.
To investigate the biological significance of Dp110 activation by Ras1, a Dp110 mutant that is unable to bind Ras1 was generated. This was achieved by mutating four residues in the Dp110 RBD (T231D, K250A, R253A and K257A), which are homologous to key residues at the PI(3)K–Ras interface in the crystal structure of the human p110 RBD bound to H-Ras3. This mutant, Dp110RBD, is similar to the p110DASAA mutant, which does not bind Ras but has normal basal PI(3)K activity in vitro3. Ras1 and Ras2 are highly conserved, particularly in the switch regions that interact with PI(3)K. Thus, although it is unlikely that Ras2 activates Dp110, our experimental approach should also disrupt any interaction between Ras2 and Dp110. In addition, although it is formally possible that the RBD mediates interactions with other proteins and that mutating it disrupts this, no such interactions have been described to date.
The inability of Dp110RBD to bind directly to Ras1 was established in a yeast two-hybrid assay (see Supplementary Information, Fig. S2a). Pulldown assays from lysates of Drosophila S2 cells expressing epitope-tagged wild-type Dp110 or Dp110RBD and p60 showed that the RBD mutation also blocks the interaction between Ras1 (data not shown) or Ras1V12 and full-length Dp110 associated with p60 (Fig. 1a). In contrast, a dominant-negative Ras1 mutant that cannot bind effector proteins, Ras1N17, does not bind to either form of Myc–Dp110 (Fig. 1a). We next determined whether other biochemical properties of Dp110RBD are similar to those of wild-type Dp110. Dp110RBD and wild-type Dp110 coexpressed with p60 accumulate to similar levels, and bind p60 to the same extent (Fig. 1b). Finally, similar amounts of immunoprecipitated Dp110RBD and wild-type Dp110 from S2 cell lysates (data not shown) and fly lysates (Fig. 1c and see Supplementary Information, Fig. S2b) possess similar in vitro lipid kinase activity. Thus, Dp110RBD is unable to interact with Ras1, but otherwise has biochemically similar properties to wild-type Dp110.
Figure 1: Dp110RBD does not bind Ras1, but is otherwise biochemically similar to wild-type Dp110.
(a) Dominant-negative Ras1N17 does not associate with p60–Dp110 or p60–Dp110RBD in pulldown assays, whereas constitutively active Ras1V12pulls down p60–Dp110 but not p60–Dp110RBD. Results shown are representative of three independent experiments. (b) Immunoblotting shows that Myc–Dp110 and Myc–Dp110RBD accumulate to similar levels in stable S2 cell lines (lower panels) and coimmunoprecipitate with similar amounts of p60 (upper panels). Results shown are representative of three independent experiments. (c) In vitro lipid kinase assays were performed on anti-p60 and anti-Myc immunoprecipitates from control, hypomorph, Myc–Dp110 or Myc–Dp110RBD flies, and the products resolved by thin layer chromatography (TLC). There is no detectable difference in the in vitro lipid kinase activity of Myc–Dp110RBD compared with Myc–Dp110. Results shown are representative of four independent experiments. Uncropped images of the whole gel and TLC plate scans are shown in the Supplementary Information, Fig. S4a–c.
Full size image (18 KB)To test the biological function of Dp110RBD, genomic transgenes were generated to express Myc–Dp110RBD or wild-type Myc–Dp110 under the control of the endogenous Dp110 promoter (see Methods). Three independent transgenic fly lines expressing similar levels of Myc–Dp110RBD and Myc–Dp110 (Fig. 2a) were tested for their ability to rescue the lethality associated with a Dp110 deletion (Dp110A)19. Surprisingly, all three Myc–Dp110RBD insertions rescue Dp110A lethality to the same extent as Myc–Dp110 (Fig. 2b). Furthermore, the rescued flies are fertile and produce viable offspring (data not shown). Myc–Dp110RBD and Myc–Dp110 also rescue Dp110A lethality to the same extent in flies hemizygous for p60 (Fig. 2c). Thus, even when the p60 gene dosage is halved, input from the RBD is not required for the essential in vivo functions of Dp110.
Figure 2: The interaction between Dp110 and Ras1 is not essential for Drosophila viability.
(a) Immunoblotting of head lysates shows that three independent Myc–Dp110 or Myc–Dp110RBD transgenic lines contain similar levels of Myc–Dp110. Flies expressing Myc–Dp110 under the control of GMR provide a positive control. An uncropped image of the gel is shown in the Supplementary Information, Fig. S4d. (b, c) Myc–Dp110RBD completely rescues Dp110A lethality (b), even in the presence of one copy of the p60A deletion19 (c). The observed frequency:expected frequency ratios were calculated by dividing the observed number of progeny of each genotype by the number expected according to Mendelian genetics. Using ANOVA tests to compare the observed:expected ratios between Dp110+/- flies and rescued flies, P = 0.87 (b) and P = 0.38 (c). The error bars represent s.d. from the mean of four and five independent experiments, respectively.
Full size image (18 KB)To more rigorously examine the effects of abolishing the interaction between Dp110 and Ras1, fly lines homozygous for the Dp110A mutation and containing two copies of the Myc–Dp110RBD or Myc–Dp110 transgenes were established. In vitro lipid kinase assays showed that Myc–Dp110RBD, Myc–Dp110 and endogenous Dp110 immunoprecipitated from control flies possessed similar intrinsic lipid kinase activity towards phosphatidylinositide, PtdIns(4)P and PtdIns(4,5)P2 (Fig. 1c). In contrast, Dp110 from a viable Dp110 hypomorphic combination of two kinase domain mutations has very low activity (Fig. 1c).
As the PI(3)K signalling pathway regulates imaginal disc growth and fly size13, the wing area and body mass of Myc–Dp110RBD flies were examined. Myc–Dp110RBD wings are approximately 86% of the area of wild-type Dp110 wings and their body mass is reduced to approximately 69% that of wild-type Dp110 flies (Fig. 3a and see Supplementary Information, Fig. S3a, b). Given the reduction in Myc–Dp110RBD adult fly size, we examined phospho-DAkt levels in imaginal discs (which develop into the adult epidermal structures) and brains. Basal phospho-DAkt levels were slightly but not significantly reduced, whereas insulin-stimulated phospho-DAkt levels were significantly reduced (by approximately 55%) in Myc–Dp110RBD discs and brains (Fig. 3b and see Supplementary Information, Fig. S3c). Thus, although the RBD mutations do not alter the in vitro kinase activity of Dp110 and do not affect fly viability, they do prevent Dp110 from responding maximally to insulin signalling and from stimulating maximal imaginal disc growth during development.
Figure 3: Myc–Dp110RBD flies cannot maximally activate PI(3)K signalling.
![Figure 3 : Myc|[ndash]|Dp110RBD flies cannot maximally activate PI(3)K signalling.](/ncb/journal/v8/n11/thumbs/ncb1493-f3.jpg)
(a) Representative images of male wings from control, hypomorph, Myc–Dp110 and Myc–Dp110RBD flies. The scale bars represent 0.2 mm. (b) Immunoblotting shows that insulin-stimulated phospho-DAkt levels are reduced in Myc–Dp110RBD versus Myc–Dp110 discs and brains. For quantification of seven independent experiments and uncropped images of the gels, see Supplementary Information, Figs S3c and S4e, respectively. (c) The number of eggs laid on the fifth day after eclosion. Myc–Dp110RBD flies lay significantly fewer eggs than Myc–Dp110 flies or control flies (using the Kruskal Wallis test to compare control, Myc–Dp110 and Myc–Dp110RBD flies, P = 0.0003). Similar results were obtained in three independent experiments and using flies heterozygous for two other transgene insertions. (d) Representative images of ovaries from Myc–Dp110 or Myc–Dp110RBD flies. The scale bars represent 0.25 mm. (e) Immunoblotting shows that basal phospho-DAkt levels are lower in Myc–Dp110RBD ovaries than in control or Myc–Dp110 ovaries. For quantification of three independent experiments and uncropped images of the gels, see Supplementary Information, Figs S3d and S4f, respectively.
Full size image (33 KB)Other phenotypes typical of mildly reduced PI(3)K signalling include lowered female fecundity, increased longevity and resistance to starvation or oxidative stress15. Although the RBD mutation has only a weak or no effect on most of these parameters (data not shown), it has a notable impact on female fecundity. Five-day old Myc–Dp110RBD flies lay 60% less eggs than Myc–Dp110 or control flies (Fig. 3c). This reduced fecundity is not due to delayed ovary maturation — their egg-laying peaks at the same age as the controls', and they lay fewer eggs than the controls at all ages (data not shown). Consistent with their reduced fecundity, Myc–Dp110RBD ovaries are small with few ovarioles (Fig. 3d). Dp110 hypomorphs lay no eggs at all and we were unable to locate their ovaries. Egg production is regulated by insulin signalling11, 12 and requires a vast amount of growth. Thus, we hypothesise that maximal egg production requires maximal PI(3)K signalling, with input from both p60 and Ras1. Consistent with this hypothesis, basal phospho-DAkt levels, and hence PI(3)K signalling, are reduced by approximately 50% in Myc–Dp110RBD ovaries (Fig. 3e and see Supplementary Information, Fig. S3d).
To summarise, we have replaced endogenous Dp110 with a mutant version that is unable to interact with Ras1 but is otherwise biochemically normal (Fig. 1) and we find that the interaction between Dp110 and Ras1 is not essential for normal fly development (Fig. 2). Nevertheless, Myc–Dp110RBD flies are small and insulin-induced DAkt phosphorylation is reduced in Dp110RBD discs; Myc–Dp110RBD flies also lay fewer eggs and basal PI(3)K signalling is reduced in Dp110RBD ovaries (Fig. 3). Our data provide the first demonstration that the interaction of a PI(3)K with Ras is essential in specific developmental contexts requiring maximal PI(3)K signalling. This work also illustrates how different levels of signalling, required to drive different biological processes, can be achieved through multiple signalling inputs. The Dp110–Ras1 interaction is likely to be more relevant to Drosophila in the wild, where environmental conditions are less optimal and fluctuate more than in the laboratory. Given the conservation of the PI(3)K RBD in other species, input from Ras is likely to be important for PI(3)K function in higher organisms. Indeed, another situation where the ability of Ras to maximally activate PI(3)K signalling and growth may be crucial is during oncogenic Ras-driven tumorigenesis.
Note added in proof: a related manuscript by Suire et al. (Nature Cell Biol. 8, doi: 10.1038/ncb1494; 2006) is also published in this issue.
Methods
Fly stocks, genotypes and rescue experiments.
Flies were raised and crossed at 25 °C according to standard procedures. w1118 flies (Bloomington stock centre, Indiana University, Bloomington, IN), were used throughout as a wild-type control. The hypomorph flies used throughout were Dp1106M5/Dp1106M2, generated from stocks kindly provided by H. Stocker and E. Hafen (Eidenössiche Tecnische Hochschule, Zurich, Switzerland). Dp1106M5 mutates Gly 823 to Asp and Dp1106V2 mutates Leu 896 to Phe; both are weakly interfering mutations in the kinase domain. The glass multimer reporter (GMR)–Myc–Dp110 flies used as a control for Fig. 2a were recombinants generated using a GMR–GAL4 line (816) provided by M. Freeman (Laboratory of Molecular Biology, Cambridge, UK) and P[UAST–Myc–Dp110]14. All other stocks used were generated in our laboratory. Transgenic flies were generated using P-element-mediated germline transformation: w1118 embryos were injected with the appropriate genomic rescue construct.
For Fig. 2a, genotypes were as follows. Myc–Dp110 lines 1 and 3: w;P[w+, gen–Myc–Dp110] on III. Myc–Dp110 line 2: w;P[ w+, gen–Myc–Dp110] on II. Myc–Dp110RBD lines 1–3: as above but with Dp110RBD instead of Dp110. For the rescue experiments in Fig. 2b and c, crosses were set up to generate 10–50 male and female flies of each test genotype. The number of progeny of each genotype of interest was divided by the number expected according to Mendelian genetics to give the observed frequency: expected frequency ratio. For Fig. 2b, genotypes were as follows. Dp110+/-: w;Dp110A, P[genH]/+. Dp110-/-: w;Dp110A/Dp110A, P[genH]. Dp110-/- + Myc–Dp110: w;P[w+, gen–Myc–Dp110], Dp110A/Dp110A, P[genH] (lines 1 and 3) or w;P[w+, gen–Myc–Dp110]/+;Dp110A/Dp110A, P[genH] (line2). Dp110-/- + Myc–Dp110RBD: as above but with Dp110RBD instead of Dp110. For Fig. 2c, genotypes were as follows. Dp110+/-: p60A/+;P[genH], P[genR10], Dp110A/+. Dp110-/- + Myc–Dp110: p60A/+;P[genH], P[genR10], Dp110A/Dp110A, P[w+, gen–Myc–Dp110]. Dp110-/- + Myc–Dp110RBD: as above but with Dp110RBD instead of Dp110. See reference 19 for additional explanation of genotypes. The Myc–Dp110 and Myc–Dp110RBD lines used in Fig. 1c and Fig. 3 contain two copies of line 1 w;P[w+, gen–Myc–Dp110] or w;P[w+, gen–Myc–Dp110RBD], two copies of P[genH] and two copies of Dp110
.
Measuring wing areas and fly mass.
Flies were raised at a density of 70 larvae per vial. For wing area measurements, wings were dehydrated in ethanol and mounted in euparal (Agar Scientific, Stansted, UK). Images were collected using a Leica MZFLIII microscope (Leica, Wetzlar, Germany) with a Nikon DXM 12000 digital camera (Nikon, Tokyo, Japan) attached, and processed and measured using Adobe Photoshop. Higher quality images of wings were obtained using a Zeiss Axioplan 2 microscope (Zeiss, Oberkochen, Germany). For fly mass measurements, the masses of batches of 20–100 flies (three to six days after eclosion) were measured, and the mean mass per fly calculated.
Fecundity assays.
Virgin female flies were placed in individual vials with two w1118 males. Four days later, the flies were transferred into fresh vials containing a few grains of yeast and the number of eggs laid by each female over a 24-h period was counted for five hypomorphs and least 14 flies of the other genotypes.
Statistical analyses.
Where statistical analyses were performed, data were plotted and normality assessed by eye. With the exception of the fecundity data, all data seemed to be normally distributed and statistical analyses were carried out accordingly. For comparison of several groups, ANOVA was used; for pairwise comparisons of two groups, a two-tailed t-test was used. In the case of the fecundity assays, the data were not normally distributed, so a non-parametric test (the Kruskal Wallis test) was used to compare the groups.
Plasmid construction.
For yeast two-hybrid assays, the entire Ras1 coding sequence and amino acids 168–317 of Dp110 were used. The following primers were used in a high-fidelity PCR reaction with Pfu Turbo DNA polymerase (Stratagene, La Jolla, CA), to introduce an EcoR1 site at the beginning and a BamHI site at the end of the desired regions: 5'-GAATTCATGACGGAATACAAACTGGTCG-3' and 5'-GGATCCGAGCATTTTACATTTAAATCTACGATTCGGC-3' for Ras1, and 5'-GAATTCCTGTATATGACCCAAACTTGCG-3' and 5'-GGATCCCTCGTTGTGGTGGTTGATGTAGG-3' for Dp110. pGEX–Ras1V12 or pGEX–Ras1N17 (see below) were used as templates for Ras1, and pMK33–Myc–Dp110 or pMK33–Myc–Dp110RBD (see below) as templates for Dp110. The blunt-ended PCR products were ligated into pCR–Blunt II TOPO using the Zero Blunt TOPO cloning kit (Invitrogen, Carlsbad, CA). The EcoR1–BamHI fragment of the Ras1 construct was ligated into EcoR1–BamHI-cut pGBKT7 ('prey' vector from the Matchmaker Gal4 Two-Hybrid System 3; Clontech, Mountain View, CA), while the EcoR1/BamHI fragment of the Dp110 construct was ligated into EcoR1–BamHI-cut pGADT7 ('bait' vector). pMTIZ–HA–p60 was generated by inserting the XbaI–NotI fragment of pBluescript(SKII) –HA–p60 (ref. 19) into XbaI–NotI-cut pMTIZ. To generate the Dp110 plasmids, a linker sequence containing useful restriction enzyme sites was first introduced into pBluescript–Myc–Dp110 (ref. 14), immediately 5' of the coding sequence. The oligonucleotides 5'-CCTCGAGGATATCGTAC-3' and 5' CATGGGAGCTCCTATAG-3' were annealed and ligated into KpnI-cut pBluescript–Myc–Dp110 to produce pBluescript–linker–Myc–Dp110. The QuikChange site-directed mutagenesis kit (Stratagene) was used to introduce the RBD mutations into this plasmid. First the T231D mutation was introduced using the primers 5'-CGAGAACGACCAGAGCACGTTTGACTTGTCGGTGAACGAGCAGG-3' and 5'-CCTGCTCGTTCACCGACAAGTCAAACGTGCTCTGGTCGTTCTCG-3'; this was followed by mutagenesis to introduce the K250A, R253A and K257A mutations, using the primers 3'-GCACGCTGCAGGCGATGAATGCGTCGCAGATGGCAATGAACGACC-5' and 5'-GGTCGTTCATTGCCATCTGCGACGCATTCATCGCCTGCAGCGTGC-3'. To generate pMK33–Myc–Dp110 and pMK33–Myc–Dp110RBD, the wild-type or mutated versions of pBluescript–linker–Myc–Dp110 were cut with XbaI and EcoRV, and the resulting 3.3 kb fragment was inserted into EcoRV–SpeI-cut pMK33.
To generate the genomic rescue constructs, the BglII fragment of pCaSpeR–Dp110 (P[gDp110])19 was replaced with the BglII fragments from wild type or mutant pBluescript-linker–Myc–Dp110. For the GST–Ras1 plasmids, point mutations were introduced into pGEX–Ras1 (gift from M. White, UT Southwestern Medical Center, University of Texas, Dallas, TX), using the QuikChange site-directed mutagenesis kit with the following primers: 5'-GGTCGTCGTTGGAGCCGTAGGCGTGGGC-3' and 5'-GCCCACGCCTACGGCTCCAACGACGACC-3' to generate pGEX–Ras1V12, and 5'-GGCGTGGGCAAGAACGCGCTCACCATCC-3' and 5'-GGATGGTGAGCGCGTTCTTGCCCACGCC-3' to generate pGEX–Ras1N17.
Yeast two-hybrid assay.
Yeast two-hybrid assays were carried out using the Matchmaker Gal4 Two-Hybrid System 3 (Clontech), according to the manufacturer's protocol, with Ras1V12 as the prey and the wild-type or mutant RBD of Dp110 as the bait. 'Full medium' lacked Leu and Try to select for the presence of the two plasmids. 'Selective medium' also lacked adenine and His, and contained X--Gal. Yeast transformed with each plasmid individually were able to grow on medium that lacked either Leu or Try and unable to grow on medium lacking both or selective medium. We also attempted to perform the converse yeast two-hybrid assay, using Ras1 as the bait and the RBD of Dp110 as the prey. However, the Ras1 bait construct alone conferred the ability to grow on selective medium.
Cell culture and transfection.
All reagents were from Invitrogen unless otherwise stated. S2 cells were cultured in Schneider's medium supplemented with 10% foetal bovine serum (Sigma, St Louis, MO), and penicillin (50 units ml-1)–streptomycin (50 mg ml-1), and maintained at 25 °C. Polyclonal S2 cell lines stably expressing HA–p60 and Myc–Dp110 or Myc–Dp110RBD were generated by calcium phosphate cotransfection with pMTIZ and pMK33 (empty vector controls), or with pMTIZ–HA–p60 and pMK33–Myc–Dp110 or pMK33–Myc–Dp110RBD. Pools of stably transfected cells were selected using 300 
g ml-1 hygromycin.
Preparation of lysates.
S2 cells were seeded at 5 
106 cells ml-1 in six-well plates. Eight hours later, the medium was supplemented with 700 g ml-1 CuSO4 to induce protein expression from the metallothionein promoter in pMK33 and pMTIZ. Sixteen hours later, the cells were harvested, washed in PBS and lysed for 10 min at 4 °C in modified RIPA buffer26. Lysates were clarified by centrifugation at 16,000g for 10 min. To prepare fly head lysates, heads were detached and frozen on dry ice. Thirty heads were homogenised in 40 l modified RIPA buffer for each lane of a gel, and lysed for 10 min at 4 °C, followed by centrifugation. The same method was used to prepare ovary lysates, using 15 control or Myc–Dp110 fly ovaries, or 20 Myc–Dp110RBD fly ovaries, in 30 l modified RIPA buffer for each lane of a gel. Immunoprecipitations were performed according to standard procedures.
Pulldown assays.
Equivalent amounts of GST–Ras1V12 or GST–Ras1N17 beads (purified according to standard protocols) were incubated with a ten-fold molar excess of a non-hydrolysable GTP analogue, GppNHp (Sigma), in the presence of 5 mM MgCl2 and 5 mM EDTA at room temperature for 20 min. The loading reaction was stopped in 10 mM MgCl2 at 4 °C and the beads were incubated with control, Myc–Dp110 plus HA–p60 or Myc–Dp110RBD plus HA–p60 S2 cell lysates for 2 h at 4 °C. Beads were washed three times in lysis buffer and once in TBS and the bound proteins analysed by immunoblotting.
Immunoblotting.
Western blots were blocked in Odyssey blocking buffer (Li-Cor, Lincoln, NE), probed with primary antibody, washed with TBS-T (TBS containing 0.1% Tween-20), then incubated with secondary antibodies conjugated with IRDye 800 (Rockland Immunochemicals, Gilbertsville, PA) at 1:5000, or Alexa Fluor 680-conjugated antibody (Invitrogen) at 1:8000. Blots were washed then visualised using the Odyssey infrared imaging system (Li-Cor). This system has a much wider linear range than chemiluminescence and therefore allows much more accurate quantification of Western blots. For Supplementary Information Figs S3c and d, seven and three independent experiments, respectively, were quantified, relative to a standard on the same blot, using Odyssey software (Li-Cor).
Kinase assays.
Kinase assays were performed as previously described14, except for the following modifications: Dp110 was purified from flies — approximately 100 flies per assay were frozen on dry ice, and homogenised and lysed in 500 l kinase assay lysis buffer (50 mM Tris–HCl at pH 7.4, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 1 mM NaF, 1 M sodium orthovanadate, 1 mM PMSF, 15 M N-tosyl-L-lysine-chloromethyl ketone (TLCK), 10.5 g ml-1 leupeptin, 10 g ml-1 aprotinin, 10 g ml-1 pepstatin). Dp110 was immunoprecipitated from clarified lysates as described above. The lipid substrates (all from Sigma) were as follows: Phosphatidylinositol, 167 g ml-1L--phosphatidylinositol; PtdIns(4)P, 83 g ml-1L--phosphatidylinositol 4-monophosphate with 83 g ml-1 1,2-diacyl-sn-glycero-3-phospho-L-serine; PtdIns(4,5)P2, 83 g ml-1L--phosphatidylinositol 4,5-diphosphate with 83 g ml-1 1,2-diacyl-sn-glycero-3-phospho-L-serine. Mammalian GST–p110 (gift from B. Vanhaesebroeck, Ludwig Institute for Cancer Research, London, UK) was used as a positive control to confirm identification of the three phosphorylated PI(3)K products. The reaction was allowed to proceed for 1 h at 30 °C with constant agitation.
Insulin stimulation of larval brains and discs.
Mouth hooks were pulled from wandering third instar larvae, and the brain and imaginal disc tissue was left attached. Mouth hooks were incubated for 1 h in Schneider's medium (control) or Schneider's medium containing 20 M insulin (Sigma) at room temperature. After freezing on dry ice, six mouth hooks per treatment were lysed in 30 l modified RIPA buffer, and lysates processed as described above.
Antibodies.
The following antibodies were used at 1:1000 dilutions for western blotting: anti-c-Myc monoclonal (clone 9E10; Sigma), anti-c-Myc polyclonal (Santa Cruz Biotechnology, Santa Cruz, CA), anti-p60 polyclonal (p60C)18, anti--tubulin monoclonal (clone E7; Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA), anti-Dp110 polyclonal14, anti-CT-DAkt16 (which recognises all forms of DAkt) and anti-phospho-DAkt16 (which recognises DAkt phosphorylated on the hydrophobic motif, Ser 505 or Ser 586 in p66DAkt or p80DAkt, respectively). For immunoprecipitations, 5 l 9E10 or anti-p60 was used per sample.
Immunofluorescence microscopy.
Dissected late third instar wing imaginal discs were fixed in methanol-free 4% (w/v) formaldehyde (Polysciences, Warrington, PA) in PBS containing 0.1% Tween-20 (PBST) for 30–45 min, then incubated in PBST with 0.1% bovine serum albumin (PBST–BSA) for 1 h. Discs were incubated overnight in DAkt antisera diluted 1:100 in PBST–BSA then washed and stained with secondary antibodies using standard protocols. Discs were mounted in Fluoro Guard antifade reagent (Bio-Rad) and analysed using a Zeiss Axioplan 2 confocal microscope and Zeiss LSM 2.5 software.
Note: Supplementary Information is available on the Nature Cell Biology website.
