Cyclin G and the Polycomb Repressive complexes PRC1 and PR-DUB cooperate for developmental stability

Abstract

In Drosophila, ubiquitous expression of a short Cyclin G isoform generates extreme developmental noise estimated by fluctuating asymmetry (FA), providing a model to tackle developmental stability. This transcriptional cyclin interacts with chromatin regulators of the Enhancer of Trithorax and Polycomb (ETP) and Polycomb families. This led us to investigate the importance of these interactions in developmental stability. Deregulation of Cyclin G highlights an organ intrinsic control of developmental noise, linked to the ETP-interacting domain, and enhanced by mutations in genes encoding members of the Polycomb Repressive complexes PRC1 and PR-DUB. Deep-sequencing of wing imaginal discs deregulating CycG reveals that high developmental noise correlates with up-regulation of genes involved in translation and down-regulation of genes involved in energy production. Most Cyclin G direct transcriptional targets are also direct targets of PRC1 and RNAPolII in the developing wing. Altogether, our results suggest that Cyclin G, PRC1 and PR-DUB cooperate for developmental stability.

Fig 1. Local deregulation of CycG induces high FA.

Wing length FA (FA10) of females bearing a Gal4 driver either associated with UAS-CycG^ΔP (dark orange) or alone (light orange). Wing length was measured as the distance between landmarks 3 and 13. (F-tests, *** p-value<0.001, S1 Table). Source data are provided in S2 Table.

Fig 2. The ETP interacting domain limits CycG-induced FA.

A–Map of the 566 amino-acid Cyclin G protein showing the ETP interacting and PEST domains. B–Wing centroid size FA (FA10) of females da-Gal4/+ (+), +/UAS-CycG^FL; da-Gal4/+, (CycG^FL) and +/UAS-CycG^ΔE; da-Gal4, (CycG^ΔE). C–Wing centroid size FA (FA10) of females da-Gal4/+ (+), +/ UAS-CycG^ΔP; da-Gal4/+ (CycG^ΔP) and +/UAS-CycG^ΔEΔP; da-Gal4/+ (CycG^ΔEΔP). (F-tests, *** p-value<0.001, S3 Table). Source data are provided in S4 Table.

Fig 3. CycG interacts with several PcG and ETP genes for developmental stability.

Centroid size FA (FA10) of ETP or PcG heterozygous mutant females combined with da-Gal4, UAS-CycG^ΔP (dark orange; da-Gal4, UAS-CycG^ΔP; PcG/+ or da-Gal4, UAS-CycG^ΔP; ETP/+) and ETP or PcG heterozygous mutant females combined with da-Gal4 (blue; da-Gal4/+; PcG/+ or da-Gal4; ETP/+). The grey dashed line shows FA of da-Gal4, UAS-CycG^ΔP/+ females. (F-tests, *p-value<0.05; ** p-value<0.01; *** p-value<0.001, S5 Table). Source data are provided in S6 Table.

Fig 4. Cyclin G co-localizes with H2AK118ub at many sites on polytene chromosomes but overexpression of CycG does not modify global H2AK118ub.

A, A’, A”–Immunostaining of polytene chromosomes from w¹¹¹⁸ third instar larvae. H2AK118ub (red), Cyclin G (green), DAPI (blue). A”‘–Close-up of the box showed in A”. B, B’–Wing imaginal discs of 3^rd instar larvae expressing CycG^ΔP in the posterior compartment under control of the en-Gal4 driver, stained with anti-Cyclin G (green) and anti-H2AK118ub (red). C, C’–Wing imaginal discs of 3^rd instar larvae expressing CycG^ΔEΔP in the posterior compartment under control of the en-Gal4 driver, stained with anti-Cyclin G (green) and anti-H2AK118ub (red). D, D’, D”, D”’–GFP clones in wing imaginal discs stained with anti-H2AK118ub (red). D’, D” and D”’ are close-up views of the yellow rectangle shown in D. E, E’, E”, E”’–CycG^ΔP clones marked by GFP in wing imaginal discs stained with anti-H2AK118ub (red). E’, E” and E”’ are close-up views of the yellow rectangle shown in E. F, F’, F”, F”’–CycG^ΔEΔP clones marked by GFP in wing imaginal discs stained with anti-H2AK118ub (red). F’, F” and F”’ are close-up views of the yellow rectangle shown in F. Scale bars: 50 μm.

Fig 5. Genes deregulated in wing imaginal discs expressing CycG^ΔP.

A–RT-qPCR analysis of endogenous CycG expression in da-Gal4,UAS-CycG^ΔP/+ and da-Gal4/+ wing imaginal discs. Expression of CycG was normalized on the geometric mean of Lam and rin (S8 Table). t-tests, ** p-value<0.01. Error bars correspond to standard deviations. B–Ontology of up-regulated and down-regulated genes in da-Gal4, UAS-CycG^ΔP/+ vs da-Gal4/+ wing imaginal discs. Gene ontology analysis was performed with DAVID (S9 Table). C–RT-qPCR analysis of RPL15, RPL7 and Rack1 expression in da-Gal4, UAS-CycG^ΔP/+ and da-Gal4/+ wing imaginal discs. Expression of RPL15, RPL7 and Rack1 were normalized on the geometric mean of Lam and rin (S10 Table). t-tests, ** p-value<0.01. Error bars correspond to standard deviations. t-tests, ** p-value<0.01; *** p-value<0.001.

Fig 6. Identification of Cyclin G genome-wide binding sites in wing imaginal discs.

A–Repartition of feature types among the 3363 decile-ranked Cyclin G peaks (S12 Table). Validated peaks were ranked based on their height (highest number of overlapping reads), and separated in ten bins before annotation. B–Average profile of Cyclin G signal over these genes shown as an aggregation plot. The standard error is represented as a shaded area over the curve. C–ChIP-qPCR analysis of RPL7, RPL5, Rack1 and CycG. IPs were performed either with Myc antibody (α-Myc) to reveal the presence of Cyclin G, or with rabbit IgG as negative control (mock). qPCR were performed using oligonucleotide primers located either at the TSS or in the gene body as indicated. Error bars represent the coefficient of variation (CV) (S13 Table). D–Ontology of the 889 genes. Gene ontology analysis was performed with DAVID.

Fig 7. Snapshots illustrated Cyclin G ChIP-seq.

Focus on RPL7 (A), RPL5 (B), Rack1 (C) and CycG (D).

Fig 8. Cyclin G shares target genes with PRC1, Asx and RNAPolII but not with Calypso.

Venn diagrams showing the intersection between Cyclin G-bound genes in +/ UAS-CycG^ΔP; da-Gal4/+ wing imaginal discs with Pc and RNAPlII (A), Asx and Calypso (B), and K3K27me3 (C) in wild-type wing imaginal discs.

Fig 9. Functional subnetwork identified in wing imaginal discs expressing CycG^ΔP.

Schematic representation of a sub-network of 222 genes centred on Cyclin G (CycG_subnetwork.xmml) and identified using JactiveModules (Z score 48.53). In this sub-network, 65 genes were up-regulated in da-Gal4, UAS-CycG^ΔP vs da-Gal4/+ wing imaginal discs (green gradient), 124 genes were down-regulated (red gradient), and 33 genes were not significantly deregulated (grey). Genes bound by Cyclin G are circled in blue. Transcription factor genes are represented by squares. Genes were clustered depending on their function. Black edges correspond to interactions discovered in the present study. Grey edges correspond to interactions described in the literature and imported into the WID network using DroID.

Fig 10. Genes bound by Asx, Calypso, Pc, Ph, Psc, or RNAPolII, or enriched in H3K27me3 in the sub-network of 222 genes centred on Cyclin G.

Bound genes are represented in blue.