Figure 1.
Genetic screen for gcmPyx modifiers and interactions with TrxG and PcG proteins.
(A) Drawing of an adult notum. Small and large dots represent microchaetae and macrochaetae, respectively. Macrochaete symbols to the right. (B–E) Adult nota from wt (WT; B), gcmPyx/+ (C), gcmPyx/suppressor deficiency (D), gcmPyx/enhancer deficiency (E) flies. Dfs = Deficiencies. Scale bar = 200 µm. Histograms present the average number of bristles per heminotum (y-axis) in different genotypes (x-axis). In all figures, average values are indicated +/− SEM (bars); P-values from t-test are indicated in the following way: *** (P≤10−3), ** (P≤10−2), * (P≤5×10−2). Pyx stands for gcmPyx. (F) Deficiencies deleting brm, (Df(3L)brm11 and Df(3L)th102), as well as the brm mutation. P-values vs. gcmPyx/+: gcmPyx/+; Df(3L)brm11/+ (8,9×10−6); gcmPyx/+; Df(3L)th102/+ (0,02); gcmPyx/+; brm/+ (4,2×10−7). (G) gcmPyx interaction with trxG genes. P-values vs. gcmPyx/+: gcmPyx/+; brm/+ (4,2×10−7); gcmPyx/+; osa/+ (0,002); gcmPyx/UAS-osa; hsGal4/+ (0,0007); gcmPyx/+; brm/osa (3,4×10−8); gcmPyx/+; trx/+ (0,009); gcmPyx/+; ash1/+ (0,01); nej/+; gcmPyx/+ (6,9×10−7); gcmPyx/+; E(bx)/+ (1,3×10−5). (H) gcmPyx interaction with PcG genes. Color code indicates members of the same complex (dark gray: PRC1, pale gray: PRC2, black: PRC recruiter). P-values vs. gcmPyx/+: gcmPyx/+; Pc1/+ (6,2×10−6); gcmPyx/+; Pc3/+ (0,0022); gcmPyx/+; Pc15/+ (1,5×10−11); gcmPyx/+; E(z)/+ (0,001); gcmPyx/esc (2,6×10−8); gcmPyx/psq (0,03). (I) Summary of the tested TrxG and PcG mutations. From left to right: the biochemical complex, the genes within the complex, the mutant phenotype over gcmPyx (No – no effect; S – suppressor; E – enhancer) and the large deficiency phenotype (nt – gene region not covered by the kit).
Table 1.
Small deficiencies tested in the secondary screen.
Figure 2.
Pc binds to the gcm promoter region.
(A–B) Association of the gcm and gcm2 loci with PcG proteins. (A) Levels of Polycomb (Pc) binding and H3K27me3 at the gcm or gcm2 gene locus and control regions (GlacAT and Rp49) in Drosophila embryos were determined by quantitative ChIP (qChIP) experiments. Results are represented as percentage of input chromatin precipitated. The standard deviation was calculated from two independent experiments. (B) Organization of the gcm-gcm2 loci, extent of the used transgenic constructs (blue lines) and ChIP-on-chip binding profiles of indicated PcG proteins and histone marks in Drosophila embryos. Data were extracted from [33]. The plots show the ratios (fold change) of specific IP versus mock IP assays. Significantly enriched fragments (P-value<1×10−4) are shown in red. Black bars indicate the location of primers used for qChIP analysis. (C,D) Eyes from flies carrying an empty mini-w+ transgenic vector (C) or a mini-w+ vector including a 9 kb gcm transgene (D). Flies heterozygous for the transgene are on the left, homozygous ones on the right. (E–H) Polytene chromosome immuno-FISH experiments performed on the gcm locus and PcG proteins. Immuno-FISH staining in wt (w1118) flies (E,F) or flies carrying a transgene including a 9 kb region upstream of the gcm TSS (G,H), with anti-Pc (E,G) or anti-Ph (F,H) antibodies. Nuclear DAPI labeling in blue. Right panels show higher magnifications of the inserts. Double labeling (E,F) with a gcm probe (E″,F″) and anti-Pc antibody (E′″) or anti-Ph antibodies (F′″) detects colocalization (arrow) at one Pc or Ph binding site in wt; transgenic animals (G–H) show a second site of colocalization. (G″–G′″, H-Hb′″). Colocalization of gcm and Ph (arrow) in wt (D) and in the transgenic line (F).
Figure 3.
The Pc mutation rescues the gcm LOF phenotype.
(A,B) Schematic drawings showing the pupal wing at (A) 29 and (B) 16 hr APF (in all panels, anterior the top, distal to the right. Inset in (A) indicates the region shown in (C–F,I,N). L3-v, L3-1 and L3-3 indicate the sensory neurons. (C–R) Immunolabeling of 24 hr APF wings: gcm-Gal4:UAS-GFP/+ (gcm-Gal4/+), considered as wt (C–H), gcm-Gal4 (I–M) and gcm-Gal4;Pc/+ (N–R). Anti-GFP labeling (green) reflects gcm expression, anti-Repo (red) marks glia and anti-Elav (blue) marks neurons. (C–H) Bracket in (E) indicates the glial cells produced by the L3-v sensory organ precursor; bracket in (F) indicates the three proximal neurons (L3-v, ACV, E1). White arrowhead indicates the L3-v neuron. Insets indicate the regions shown at higher magnification (C,I,N). (G,H) The L3-v GP produces several GFP+/Repo+ cells (arrows). In mutant wings (I–M), the L3-v lineage produces only one GFP+ cell (J,M), which does not express Repo (K), but Elav (L,M asterisk indicates the ectopic neuron). In double gcm and Pc LOF wings (N–R), several GFP+ cells (O,R) express Repo (P) and no ectopic neurons were observed (Q). (S) Quantitative data on the fate transformation phenotype at different stages. (T–V) Immunolabeling in 9 hr APF wings: gcm-Gal4/+ (T–T″); gcm-Gal4 (U–U″) and gcm-Gal4;Pc/+ (V–V″). In all genotypes, one GFP+ cell produced by the L3-v lineage is visible (T,U,V). In the heterozygous wing, this cell expresses Repo (T′) and not Elav (T″). In gcm-Gal4 (U), the GFP+ cell does not express Repo (U′), but expresses Elav (U″). In the double gcm and Pc LOF wing, the GFP+ cell (V, empty arrowhead) expresses Repo (V′) and Elav (V′). Scale bars: C–F,I,N = 100 µm; G,H,J–M,O–R,U–W″ = 10 µm.
Figure 4.
PcG genes control glia proliferation.
(A–E) Quantitative analysis of glia at L3 vein position. Graphs comparing animals of different genotypes for the number of glia present on the L3 vein by 20 hr APF. The Y-axis indicates the percentage of wings showing a given number of glia; the X-axis, the number of glia expressing the Repo protein. Color-code is used to distinguish the compared genotypes: (A) wt vs. gcm-Gal4/+, (B) wt vs. gcm-Gal4, (C) gcm-Gal/+ vs. gcm-Gal4, (D) gcm-Gal4 vs. gcm-Gal4; Pc/+. (A) In wild type wings, the number oscillates between 5 and 7, whereas in gcm-Gal4/+ wings it oscillates between 3 and 9 cells. (B,C) gcm-Gal4 homozygous animals carry fewer Repo labeled cells and less variation (from 3 to 6) than heterozygous animals (from 3 to 9). This is also reflected by the presence of one peak value for homozygous animals and two for heterozygous animals. (D) Note that gcm-Gal4; Pc/+ animals show an increase of glial cell number compared to that observed in gcm-Gal4 animals. (E) The graph shows the distribution around the average of the number of Repo+ cells in the different genotypes as indicated by the color code. (F–H) Quantitative analysis of glia at L1 vein position. Graphs comparing animals of different genotypes for the number of glia present on the L1 vein by 24 hr APF. The Y-axis indicates the percentage of wings showing a given number of glia; the X-axis, the quantitative range of Repo expressing cells. Color-code is used to distinguish the compared genotypes: (F) wt vs. Pc/+ or vs. E(z)/+, (G) Pc/+ vs. E(z)/+ or vs. Pc/E(z). (F) Most wild type animals show from 50 to 60 glia. (G) Note that most Pc/E(z) double heterozygous animals show higher number of glia (from 70 to 80 Repo+ cells) when compared to single heterozygous animals. This is confirmed by more than 20% of wings showing over one hundred Repo+ cells on L1 vein. (H) The graph shows the distribution around the average of the number of Repo+ glia at the L1 vein position in the different genotypes as indicated by the color-code. (I) Quantitative analysis of pupal wings showing a double Repo/PH3+ cell indicating glia proliferation.
Figure 5.
gcm is overexpressed in Pc mutants.
(A–J) In situ hybridization with a gcm-specific probe. (A,B,D,E) 19 hr APF wings: gcm is expressed at the L1 nerve position (L1) and in the so-called twin sensilla of the margin (TSM) in wt (A) as well as in Pc/+ animals (D); by 24 hr APF, gcm is no more expressed in wt (B), but persists in Pc/+ wings (E) (asterisk indicates a non-specific signal). (C,F) gcm expression in the embryonic brain (arrowhead) and in the ventral cord (brackets) fades by stage 14 in wt (C), but persists in Pc mutants (F) (lateral views, anterior to the left). (G–J) optic lobe partial projection (anterior to the top; scale bar = 100 µm): in wt (G), gcm is expressed at the position of the lamina glial cell precursor (GPC) area (arrows); gcm expression in Pc/+ (H), in brm/+ (I) and in brm, trx/+ double mutants (J). Note that we focused on early third instar larvae, when the first burst of expression takes place in the GPC region. At that time, gcm is just starting being expressed in the other territories that have been previously described as gcm positive [39], [40], [43]. (K) Schematic representation of optic globe gcm-dependent lamina glial lineages. In blue, the GPCs. In green, differentiating and migrating glial cells (direction shown by the arrows). (L) Schematic representation of the areas of gcm expression (red) in the GPC region, based on the above in situ analyses.
Figure 6.
Pc inhibits gcm autoregulation and glial differentiation.
(A–I) Immunolabeling of gcm GOF embryos carrying rA87, a lacZ insertion that detects endogenous gcm expression, UAS-gcm and the scabrous-Gal4 driver, active in the whole embryonic ventral cord (white brackets) (A,D,G); gcm GOF, Pc LOF embryo (B,E,H); gcm GOF, Pc GOF embryo (C,F,I). Ventral views. gcm GOF causes endogenous gcm overexpression (A), and ectopic glial cell production (D,G). In a Pc LOF embryo, the number of ß-Gal+ cells increases (B), as well as the number of Repo+ (E,H); in a Pc GOF embryo, the number of ß-Gal+ (C) and Repo+ (F) cells decreases. (J) Histograms present the average number of ß-Gal+ (red) and Repo+ (green) cells in embryonic thoracic segments (y-axis) from different genotypes (x-axis). P-values of t- test vs. gcm GOF: gcm GOF Pc LOF (ß-Gal 1,7×10−5; Repo 1,5×10−5); gcm GOF Pc GOF (ß-Gal 0,003; Repo 0,0001). Scale bar = 100 µm. The graph (K) shows the activation of a 2 kb gcm2 promoter reporter construct displaying four GBSs (L). The ratio between reporter activity upon Gcm/Pc coexpression vs. that observed upon Gcm expression alone indicates that the gcm2 promoter is activated when Gcm is expressed in S2 cells and repressed upon Gcm and Pc coexpression.
Figure 7.
Schematic models for Pc mode of action.
(A–C) Possible mode of action of Pc on HOX genes and on a developmental gene that is transiently expressed. See the different expression profiles on the schematic graph (D). (A) promoters that are constantly active (HOX) are devoid of Pc binding. (B) Transient activator(s) may compete with PcG proteins for binding thereby modulating the levels of expression of dynamically expressed promoters. (C) Active, dynamically expressed, promoters may be constantly occupied by PcG proteins and their expression levels depend on the amount of transient activator(s) available and bound to the promoter. Color code legend is included.