Orian et al. 10.1073/pnas.0707418104.

Supporting Information

Files in this Data Supplement:

SI Figure 6
SI Figure 7
SI Table 2
SI Text




Fig. 5. dMyc and Gro shared targets and complex formation. (A) The overlap in dMyc and Gro shared targets is not due to random occurrence. Graph representing the Chance probability, P, vs. the number of overlapping dMyc-Gro shared targets identified. (B and C) Immunoprecipitation of dMyc-Gro complexes from Kc cells. Kc cell extracts were immunoprecipitated with the indicated antibodies and analyzed by Western blot. Note that the dMyc and Gro proteins are 717 and 719 aa, respectively, and therefore comigrate at »90 kDa on protein gels. GFP and cadherin IPs are nonrelevant antibody controls. For all cases, 5% input is shown. (D) Displacement of 35S-dMyc bHLHZip from GST-Gro by increasing amounts of an IVT- dMax-DAM fusion protein, but not by IVT-DAM protein alone. (E) IVT-Gro cannot displace 35S-dMax bound to GST-dMyc (506-717 aa). (F) Diagram of the dMyc protein fragments used to map domains interacting with Gro. (G) Full-length dMyc (1-717 aa), as well as proteins containing either the N-terminal (50-240 aa) or the bHLHz (624-717 aa) regions, are independently capable of binding to Gro (summarized in F). (H) Diagram of the Gro protein fragments used to map domains interacting with dMyc. Gro/TLE proteins share a similar overall domain structure, including an N-terminal glutamine-rich Q domain required for protein oligomerization, internal Ser/Thr/Pro-rich sequences, and C-terminal tandem "transducin-like" repeats (WD domain) implicated in protein-protein recognition (Fig. 4F) (ref. 1; reviewed in refs. 2 and 3). (I) IVT-35S-Met labeled full length Gro interacts with GST-Myc-NT (46-507 aa), GST-Myc-CT (506-717 aa), and Hairy. An IVT-[35]S-Met labeled Q-domain containing fragment of Gro interacts with both GST-Myc-NT and GST-Myc-CT, whereas the Gro protein lacking the Q domain binds to GST-Hairy.

1. Jiménez G, Paroush Z, Ish-Horowicz D (1997) Genes Dev 11:3072-3082.

2. Parkhurst SM (1998) Trends Genet 14:130-132.

3. Courey AJ, Jia S (2001) Genes Dev 15:2786-2796.





Fig. 6. dMyc and Gro shared targets. (A) Decline in target gene expression due to reduced dMyc levels can be partially restored by the simultaneous reduction of Gro levels. Immunostaining for Nop60B (red) in Drosophila Kc cells treated with RNAi to the indicated gene(s). Insets show a higher magnification view of one cell. DNA is identified using Dapi (blue). (B and B") Similar to the mitotic effect observed in embryos, we and others (1) have noted that reducing dMyc levels in Drosophila cells results in a binucleated phenotype, a phenotype generally associated with mitotic defects and unresolved cytokinesis (see SI Text).

  1. Eggert US, Kiger AA, Richter C, Perlman ZE, Perrimon N, et al. (2004) PLoS Biol. 2: e379.




Fig. 7. Regulation of sensory bristle (PNS) formation. (A-I) Scanning electron micrographs of adult female thoraces from wild-type (A) or from females harboring dMyc somatic clones resulting in regions with bristle loss (FRT-dMycdm4; clones are indicated by arrows) (B), females harboring Gro somatic clones resulting in regions with ectopic bristle expression (FRT-groE47; clones are indicated by arrows) (C), females expressing two copies of dMyc under the control of the C253-Gal4 driver (UAS-dMyc; note that this expression results in formation of extra bristles and partial transformation of these from microchaetes into macrochaetes) (D), females expressing one copy of Gro under the control of the C253-Gal4 driver (UAS-Gro; this expression results in reduced bristle number) (E), and females coexpressing one copy each of dMyc and Gro, resulting in a bristle phenotype similar to wild type (F). Similar to Myc, expression of activated EGFR (UAS-lTOP) results in extra sensory bristle formation (G). In contrast, ectopic expression of either dMnt (H) or dMax (I) does not affect sensory bristle formation or patterning. (J-L) dMycdm1 suppresses the formation of ectopic bristle observed in heterozygous N55e11 females. Scanning electron micrographs of adult female thoraces from dMycdm1/+ (J), N55e11/+ (K), and N55e11/dMycdm1 doubly heterozygous (K) females. The average number of bristles per 900-micron square area is indicated in the lower left-hand corner of each image. (M and N) Overexpression of dMyc results in bristle duplication. Scanning electron micrographs of two different adult female thoraces expressing two copies of dMyc under the control of the C253-Gal4 driver. Inset (M') is a magnification of the rectangle in M.





Table 2. dMyc and Groucho shared direct targets

Gene

Cytological

location

Biological process

cycA

68E1

Mitosis

cycB1

59B2

Mitosis

cycB3

96B1

Mitosis

aurora

87A3

Mitosis

aurora2/IAL

32B1-2

Mitosis

fzy

35F8-9

Mitosis

ncd

99C1-2

Mitosis

bub1

42A-1

Mitosis

fzy-related

4C11-12

Mitosis

barren

38B1-2

Mitosis

CG7783/smc5

78D6-7

Mitosis

CG31916

25C7-10

Mitosis

*Gamma-Tubulin

37C

Mitosis

Nop60B

60C1-2

Mitosis/ribogenesis

Kr

60F5

Transcription factor

ci

101A1-3

Transcription factor

vnd

1B10

Transcription factor

ind

71B2

Transcription factor

dMax

76A3

Transcription factor

CG14005

25F5

Transcription factor

CG3281

87A3

Transcription factor

cta

NA

Signaling

reaper-L

75C6

Cell death

Fpps

47F1

Cholesterol metabolism

GART

27D4

Cuticle protein

CG1890

1002D

Chaperonin (for tubulin)

CG6203 (fmr1)

85F10-12

RNAi; component of RISC complex

CG5515

95E6

Unknown

CG3165

23C4

Unknown

CG15233

42C1

Unknown

CG8710

44B3-4

Unknown

CG12424

51D6

Unknown

CG15675

57F7

Unknown

CG7049

61B2

Unknown

CG9705

73C4

Unknown

CG5969

77C4

Unknown

CG6425

97C5

Unknown

LD13080

NA

Unknown

*Based on experimental finding (not DamID).





SI Text

Scanning electron microscopy and PNS phenotypes.

Adult flies for scanning electron microscopy analysis were fixed in SH media (1 part glycerol: 7 parts ethanol) for >1 day. The adult flies were then dehydrated in an ethanol series then placed in hexamethyldisilazane (HMDS; Ted Pella). The HMDS was allowed to evaporate then the flies were mounted on carbon black tape and sputter-coated using gold-palladium. The bristle phenotypes exhibited in the individual over-expression cases (shown in SI Fig. 7) are dose sensitive, with one copy showing a weaker version of the phenotype exhibited by two copies. Regardless of the severity of the phenotype, the penetrance of the phenotypes in these individual over-expression cases (shown in SI Fig. 7 D-J), as well as the LOF cases (shown in SI Fig. 7 B and C), is >95%. The rescue of the UAS-Gro phenotype by the addition of UAS-dMyc is also dose-dependent (10% rescue with one copy and 67% rescue with two copies).

DamID chromatin profiling and network comparisons.

We previously used a microarray-based genomic chromatin profiling method termed DamID to identify the direct binding sites of the Drosophila Myc/Max/Mnt network (1). DamID, similar to the ChIP-chip chromatin profiling technique, is a powerful tool that allows systematic and global identification of in vivo direct targets of transcriptional networks (reviewed in refs. 2-4). In these cases, we identified targets for these proteins in Drosophila Kc cells or embryos using microarray chips containing »6,200 full-length Drosophila Gene Collection (DGC) cDNAs and ESTs representing roughly one-half of the fly genome. Interestingly, our genomic mapping revealed a more complex regulatory landscape for the Myc network than previously predicted. Our results showed a broad association of the Myc network with the genome (binding, »15% of coding regions) consistent with its widespread regulation of gene expression. Interestingly, although recruitment of dMyc-Dam to its genomic loci depended on dMax levels, dMnt recruitment was independent of dMax levels (1). We had also used DamID to identify direct transcriptional targets of the Drosophila Hairy bHLH transcriptional repressor and its associated cofactors, Gro, dCtBP, and dSir2 (5). In addition to finding putative targets for Hairy in segmentation, we identified groups of targets suggesting roles for Hairy in cell cycle/growth and morphogenesis, processes that must be coordinately regulated with pattern formation. Our studies also provided a global view of cofactor recruitment requirements and showed that their recruitment is highly context-dependent.

In addition to identifying direct targets for individual developmental networks, the generation of DamID data sets provides the potential for comparing the direct targets of one transcriptional or signaling network with others. Such analyses have the potential to identify unknown molecular junctions between or among networks. Using genome-wide comparison of direct transcriptional targets of the dMyc and Hairy networks, we identified a protein-protein interaction between the Drosophila Myc oncogene and the Groucho co-repressor that regulates a subset of non-canonical direct dMyc targets.

Recruitment of dMyc and Gro to target gene regulatory regions using DamID Southern analysis.

dMyc-Gro antagonism may be molecularly mediated via a single protein complex bound to one DNA site or, alternatively, via multiple independent binding sites, that may anchor a single protein complex or even several independent protein complexes (e.g., a Myc activation complex and a Gro repressive complex). Therefore, we examined the recruitment of the two proteins to genomic loci of a subset of shared targets. DamID Southern analysis was performed as described (6, 7) to perform rough mapping of the binding of dMyc and Gro along the genomic region of a given shared target gene (Fig. 4 I-J). Drosophila Kc cells were transfected with either Dam-dMyc or Gro-Dam and genomic DNA was isolated and cut with DpnI. Subsequently 0.2-2.0-kb fragments were isolated via a sucrose gradient and equal amounts were used for a Southern blot. Probes spanning the promoter regions ("promoter") or the coding regions ("gene") of the Cyc-B, dMax, Nop60B, and g-tubulin shared targets were used for this analysis. Although Gro appears to associate with 5' upstream regions (large region of hybridization to the higher-molecular-weight fragments in Gro lane that is absent from Myc lane), dMyc appears to associate with regions within the coding regions (discrete bands that overlap minimally between the Myc and Gro lanes).

Plasmids and constructs.

dMyc, dMyc derivatives, dMax, dMax-Dam, Dam only (8), and pSC-6MT-Gro (8) were used for in vitro translation using a TNT transcription-translation coupled kit (Promega). GST-dMyc (46-507 aa), GST-dMyc (bHLHz; 506-717 aa), GST-c-MycN262, GST-c-MycC92, GST-dMax, GST-dMnt, GST-Gro, GST-Hairy, and CMV-Max were as described (8-12). His-6-Gro was generated by subcloning a PCR derived full-length Gro ORF as a 5' BamHI-3' SalI fragment into pRSETA (Stratagene) and expressed as described (10).

Gel filtration.

Gel filtration was performed by using 20-mg/ml embryo extract (gift of T. Tsukiyama, Fred Hutchinson Cancer Research Center) loaded onto a HiLoad Superdex 200 HR gel-filtration chromatography column, and proteins were resolved in a buffer containing 20 mM Tris-HCl, pH 7.2/150 mM NaCl/2 mM DTT using an FPLC system (Pharmacia).

Northern analysis.

RNA was made as described (9). Northern production and hybridization were as described (9), using 10 mg of total RNA per lane and Magnagraph membrane (Micro Separations, Inc.). actin5C is expressed ubiquitously and was used as a loading control.

Immunoprecipitations (IPs) and immunofluorescence.

We used antibodies to the following proteins for both the IPs and Western blotting (WB): a-Gro at 1:10 (IP) and 1:400 (WB); a-dMyc p4c4b10 IgG1 at 1:2 (IP) and 1:50 (WB); a-dMnt at 1:10 (IP) and 1:100 (WB); a-DE-Cadherin at 1:15 (IP) and 1:10 (WB); and a-GFP at 1:60 (IP). We used antibodies to the following proteins for immunostaining(s) and/or Western blots (b): a-Gro at 1:40 (s) and 1:300 (l) (from C. Delidakis, IMBB, Crete); a-dMyc p4c4b10 IgG1 at 1:3 (s) and 1:100 (b); a-Hb at 1:300 (s) and a-Kr 1:200 (s) (from Jon Reinitz, University at Stony Brook); a-vnd at 1:2000 (s) (from D. Mellerick, University of Michigan); a-pSer10-Histone H3 at 1:1000 (s) (Upstate Biotechnology); a-Nop60B at 1:250 (s) and 1:1000 (b) (from S. Poole, University of California, Santa Barbara); a-g Tubulin at 1:200 (b) (from S. Henikoff, Fred Hutchinson Cancer Research Center); a-CycA at 1:500 (b) and a-CycB at 1:2,000 (b) (from D. Glover, University of Dundee); a-Barr at 1:4000 (b) (from H. Bellen, Baylor College of Medicine); and a-22C10 at 1:20 (s) (Developmental Studies Hybridoma Bank). Quantification was performed using an Ag-ECL machine or ImageJ software.

Drosophila

cell culture and RNAi experiments.
Kc167 cells were maintained at 25°C in CCM3 serum-free media (HyClone) with 10 mM glutamine. Cells completely lacking dMyc will not cross the G1/S checkpoint, and the binucleated phenotype we observe in dMyc RNAi-treated cells is similar to that observed by Eggert et al. (13) for cell cycle regulators such as cyclin-A. However, this binucleate phenotype would also be consistent with cells that are unable to undergo cytokinesis. The RNAi in our experiments was applied to an asynchronous population of cells, and it is possible that the cell cycle phase in which dMyc is silenced affects the cell phenotype (i.e., cells that have not yet entered S phase will not cross the G1/S checkpoint, whereas cells that have already entered S phase might retain enough cyclin-A and -B stores to advance to mitosis). If such cells lacking dMyc advance, they would likely fail at this stage and not complete cytokinesis.

Primers used to silence genes in Drosophila cells.

dMyc:

Forward: 5'-GAATTAATACGACTCACTATAGGGAGAATCAACAGCATGTCGCAACA-3'

Reverse: 5'-GAATTAATACGACTCACTATAGGGAGAGAATCCACTGTGTTGCGTCCA-3'

Gro:

Forward: 5'-TTAATACGACTCACTATAGGGAGAGATCCACGCCCAGCAG-3'

Reverse: 5'-TTAATACGACTCACTATAGGGAGTTCATCCTCTTGCAGTT-3'

GFP:

Forward: 5'-GAATTAATACGACTCACTATAGGGAGACTACCTGGTTCCATGGCCAAC-3'

Reverse: 5'- GAATTAATACGACTCACTATAGGGAGAAAAGGGCAGATTGTGTGGAC-3'

E-box-binding site bioinformatics analysis.

Promoter sequences of the Myc-Gro shared targets were obtained by using Apollo (14). Promoter sequences were screened for the presence of E-box (CACGTG) by using the RSA DNA Pattern Tool (15). The number of E-box occurrences ranged from 0 to 6 per promoter.

1. Orian A, van Steensel B, Delrow J, Bussemaker HJ, Li L, Sawado T, Williams E, Loo LW, Cowley SM, Yost C, et al. (2003) Genes Dev 17:1101-1114.

2. Orian A (2006) Curr Opin Genet Dev 16:157-164.

3. Greil F, Moorman C, van Steensel B (2006) Methods Enzymol 410:342-359.

4. Van Steensel B (2005) Nat Genet 37:S18-S24.

5. Bianchi-Frias D, Orian A, Delrow JJ, Vazquez J, Rosales-Nieves AE, Parkhurst SM. (2004) PLoS Biol 2:E178.

6. Grewal SS, Li L, Orian A, Eisenman RN, Edgar BA. (2005) Nat Cell Biol 7:295-302.

7. van Steensel B, Henikoff S. (2000) Nat Biotechnol 18:424-428.

8. Poortinga G, Watanabe M, Parkhurst SM (1998) EMBO J 17:2067-2078.

9. Phippen TM, Sweigart AL, Moniwa M, Krumm A, Davie JR, Parkhurst SM (2000) J Biol Chem 275:37628-37637.

10. Rosales-Nieves AE, Johndrow JE, Keller LC, Magie CR, Pinto-Santini D, Parkhurst SM (2006) Nat Cell Biol 8:367-376.

11. Gallant P, Shiio Y, Cheng PF, Parkhurst SM, Eisenman RN (1996) Science 274:1523-1527.

12. Loo LW, Secombe J, Little JT, Carlos LS, Yost C, Cheng PF, Flynn EM, Edgar BA, Eisenman RN (2005) Mol Cell Biol 25:7078-7091.

13. Eggert US, Kiger AA, Richter C, Perlman ZE, Perrimon N, et al. (2004) PLoS Biol. 2: e379.

14. Lewis SE, Searle SM, Harris N, Gibson M, Lyer V, et al. (2002) Genome Biol 3:research0082.

15. van Helden J (2003) Nucleic Acids Res 31:3593-3596.