SCFSlimb ubiquitin ligase suppresses condensin II-mediated nuclear reorganization by degrading Cap-H2

Abstract

Condensin complexes play vital roles in chromosome condensation during mitosis and meiosis. Condensin II uniquely localizes to chromatin throughout the cell cycle and, in addition to its mitotic duties, modulates chromosome organization and gene expression during interphase. Mitotic condensin activity is regulated by phosphorylation, but mechanisms that regulate condensin II during interphase are unclear. Here, we report that condensin II is inactivated when its subunit Cap-H2 is targeted for degradation by the SCF(Slimb) ubiquitin ligase complex and that disruption of this process dramatically changed interphase chromatin organization. Inhibition of SCF(Slimb) function reorganized interphase chromosomes into dense, compact domains and disrupted homologue pairing in both cultured Drosophila cells and in vivo, but these effects were rescued by condensin II inactivation. Furthermore, Cap-H2 stabilization distorted nuclear envelopes and dispersed Cid/CENP-A on interphase chromosomes. Therefore, SCF(Slimb)-mediated down-regulation of condensin II is required to maintain proper organization and morphology of the interphase nucleus.

Figure 1.

SCF^Slimb RNAi promotes interphase chromatin compaction. (A–D) 7-d RNAi-treated S2 cells stained with Hoechst to visualize DNA. Depletion of Cul-1 (B), SkpA (C), or Slimb (D) but not control (A) promotes interphase chromatin compaction, generating a “gumball” phenotype. Cells are shown at low and high magnifications (left and middle). Shown on the right are 3D surface plots of the fluorescence intensities of the DNA (insets). (E) Representative images of DNA-stained RNAi-treated S2 cells displaying normal (wild-type), weak gumball, and strong gumball phenotypes. (F) Frequency histogram of the nuclear phenotypes in S2 cells after a 7-d depletion of the indicated proteins (n = 1,400–1,800 cells per treatment). (G) Chromatin of S-phase arrested cell compacts after slimb RNAi. S2 cells were treated daily with DMSO or S-phase arrested with hydroxyurea + aphidicolin for 6 d. Beginning on day 2, cells were also treated daily with control or slimb RNAi (see Fig. S2 A). Histogram shows the frequencies of nuclear phenotypes on day 6 (n = 1,100–1,600 cells per treatment). (H) S2 cells restricted to interphase form compact chromatin domains after slimb RNAi. Cells were treated daily with control, String, or cyclin A (cycA) dsRNA for 8 d. Stg RNAi promotes G2 arrest whereas cycA RNAi blocks mitotic entry. Beginning on day 4, cells were also treated daily with slimb RNAi (see Fig. S2 B). Histogram shows the frequencies of nuclear phenotypes on day 8 (n = 600–1,300 cells per treatment). Error bars indicate SEM.

Figure 2.

SCF^Slimb depletion causes chromatin reorganization, abnormal dispersal of centromeric protein Cid, and nuclear envelope defects. (A) Representative image of a Slimb-depleted DNA-stained S2 cell displaying a strong chromatin-gumball phenotype. (A′) Segmented, pseudo-colored representation of the nucleus in A. (B) cul-1, skpA, or slimb RNAi cause interphase nuclei to reorganize their chromatin into 9–13 “gumball-like” compact globules. The number of globular segments per nucleus was measured from images as in A′ (n = 100–140 cells per histogram). (C) Control and Slimb-depleted Kc cells stained for DNA (blue) and two different euchromatic FISH probes specific for the X chromosome (green and red). FISH labeling shows paired homologous X chromosomes in controls but unpaired after slimb RNAi, decorating distinct chromatin domains. (D and E) Slimb depletion induces abnormal Cid dispersal. Shown are control and slimb RNAi-treated S2 cells immunostained for Cid (red). Hoechst-stained DNA, green. Boxes in E show abnormal Cid dispersal at higher magnification (labeled 1–3 on the right). (F) slimb RNAi increases the number of Cid spots per chromatin gumball. Cid spots were measured from 285 chromatin gumballs in 36 S2 cells. Numbers above the bars are the number of gumballs in each category. (G–I) slimb RNAi causes nuclear envelope defects. Slimb-depleted S2 cells immunostained for nuclear lamin (red; grayscale in bottom panels) show envelope invaginations (G), deformed, crumpled envelopes (H), and internalized nuclear microspheres (I, arrowheads). DNA, blue. Bars: (C–E and G–I) 2.5 µm; (E, insets) 0.25 µm.

Figure 3.

Depletion of condensin II subunits rescues slimb RNAi-induced nuclear phenotypes. (A–H) 7-d RNAi-treated S2 cells stained for DNA. Whereas slimb RNAi (B) promotes chromatin compaction, double RNAi of slimb and SMC-2 (D), cap-H2 (F), or cap-D3 (H) rescues this phenotype. Control (A), SMC-2 (C), cap-H2 (E), and cap-D3 (G) single RNAi-treated cells are shown at low and high magnifications (left and middle). Right, 3D surface plots of DNA fluorescence intensities. (I) Frequency histogram of nuclear phenotypes after day 7 RNAi (n = 1,200–1,500 cells per treatment). Error bars indicate SEM. (J) Day 7 double cap-H2/slimb RNAi rescues increase in Cid numbers. S2 cells immunostained for Cid (red). DNA, green. (K) Double cap-H2/slimb RNAi prevents an increase in Cid spots. The number of Cid spots per nucleus was counted from RNAi-treated interphase cells (100 cells per histogram).

Figure 4.

Cap-H2 is degraded in a Slimb-dependent manner and stabilized by perturbing its interaction with Slimb. (A) An inducible Cap-H2-EGFP stable S2 cell line was control or slimb RNAi-treated for 7 d, and lysates were immunoblotted for the indicated proteins. Slimb depletion stabilizes Cap-H2-EGFP (and Armadillo) but not SMC-2. α-Tubulin, loading control. (B and C) Reciprocal immunoprecipitations and Western blots show that Slimb associates with Cap-H2. (B) Anti-GFP immunoprecipitates from cell lysates expressing Cap-H2-EGFP (top) or EGFP (bottom) probed with anti-GFP, Slimb, and SMC antibodies. (C) Anti-Slimb immunoprecipitates from cell lysates expressing Cap-H2-EGFP (top) or EGFP (bottom) probed for GFP and Slimb. (D) Cap-H2 is ubiquitinated. Anti-GFP immunoprecipitates from cell lysates transiently expressing 3×FLAG-ubiquitin and either Cap-H2-EGFP or EGFP probed for FLAG and GFP. (E) A Slimb-binding consensus (gray boxes) is encoded near the Cap-H2 carboxy terminus and flanked by multiple Ser/Thr residues (bold). A Cap-H2 Slimb-binding mutant (SBM) was engineered by substituting nonphosphorylatable Ala residues for two Ser residues (asterisks) within the Slimb-binding motif. Cap-H2-ΔC23-EGFP lacks these 23 terminal residues. (F) Anti-GFP immunoblots of lysates from cells transiently expressing Cap-H2-EGFP, Cap-H2-SBM-EGFP, or Cap-H2-ΔC23-EGFP. α-Tubulin, loading control. (G) Deletion of the Slimb-binding region disrupts Cap-H2 association with Slimb. Anti-GFP immunoprecipitates from cell lysates transiently expressing either inducible Cap-H2-EGFP or Cap-H2-ΔC23-EGFP probed for GFP and endogenous Slimb. (H) Cap-H2 is phosphorylated. Anti-GFP immunoprecipitates from day 7 slimb RNAi cells expressing inducible Cap-H2-EGFP were mock- or lambda phosphatase–treated. Compared with mock treatment (broken line), lambda phosphatase treatment alters the mobility of Cap-H2-EGFP to a faster migrating species (dotted line).

Figure 5.

Stable Cap-H2 promotes interphase chromatin compaction, Cid dispersal, and nuclear envelope defects. (A) DNA-stained S2 cells displaying representative nuclear phenotypes after overexpression of Cap-H2-ΔC23-EGFP. (B) Expression of stable Cap-H2 mutants induces chromatin compaction. Histogram of nuclear phenotype frequencies in cells expressing inducible EGFP, Cap-H2-EGFP, or Cap-H2-ΔC23-EGFP for 24 h. Expression levels were modulated by inducing with either 0.5 or 1 mM CuSO₄ (n = 1,200–1,800 cells per treatment). Error bars indicate SEM. (C) Expression of stable Cap-H2 promotes abnormal Cid dispersal. S2 cells expressing EGFP (top; green) or Cap-H2-ΔC23-EGFP (bottom; green) and immunostained for Cid (red). Hoechst-stained DNA, blue. Boxed regions show abnormal Cid dispersal (insets, shown in higher magnification). (D) Stable Cap-H2 increases number of Cid spots in interphase cells. Numbers of Cid spots per nuclei were counted from interphase cells transiently expressing EGFP (top) or Cap-H2-ΔC23-EGFP (bottom; 51 cells per histogram). (E–G) Expression of stable Cap-H2 causes nuclear envelope defects. S2 cells expressing EGFP (E; green) or Cap-H2-ΔC23-EGFP (F and G; green) and immunostained for nuclear lamin (red; grayscale on the right). DNA, blue. Stable Cap-H2 expression induces invaginations and internalized nuclear microspheres (F) and crumpled envelopes (G). Bars, 2.5 µm.

Figure 6.

Slimb regulates chromatin organization in vivo, and its depletion causes unpairing of salivary gland polytene chromosomes. (A and B) Wild-type and FLP/FRT generated homozygous slimb^UU11 null mutant clone of imaginal wing disk cells immunolabeled for lamin B (red). DNA, blue. Loss of Slimb induces chromatin gumballs. Insets show single nuclei (arrowheads) at higher magnification. (C–F) After heat shock, a transgenic line carrying Hs>LacI-GFP, LacO(250): Hs>Gal4 expresses LacI-GFP, which binds LacO arrays inserted on the second chromosome. Nuclei of salivary gland cells showing DNA (grayscale and blue) and organization of LacO arrays (green). UAS>Dicer2 expression does not disassemble polytenes, so LacI-GFP localizes to a single stripe (C). In contrast, Cap-H2 overexpression (D) or Slimb depletion (E) unpairs polytenes, causing LacI-GFP spots to disperse. Weak polytene unpairing is observed in cells possessing the UAS>slimb RNAi transgene without Dicer2, so some nuclei have two or three LacI-GFP spots (F). Error bars indicate SEM. (G) In vivo Cap-H2 overexpression or slimb RNAi causes polytene chromosomes to unpair. UAS>Dicer2 control salivary glands have one large LacI-GFP spot (n = 24 nuclei), whereas UAS>slimb-RNAi larvae display a small but significant increase in unpairing (*, P < 0.005; n = 24 nuclei). Polytene unpairing is sharply increased in both UAS>Cap-H2^EY09979 and UAS>slimb-RNAi, UAS>Dicer2 lines (**, P < 10⁻¹² and **, P < 10⁻¹⁶, respectively, compared with UAS>Dicer2 alone; n = 24 nuclei per treatment). (H and I) Salivary gland nuclei expressing Hs>Gal4 (H) or Hs>Gal4, UAS>Cap-H2^EY09979 (I) stained for nuclear envelopes (wheat germ agglutinin, red). DNA, green. Bars: (A and B) 5 µm; (A and B, insets) 0.2 µm; (C–F, H, and I) 10 µm.

Figure 7.

Slimb mutations suppress condensin II loss-of-function phenotypes in polyploid nurse cells. (A) Stage 10 nurse cells from double heterozygous Smc4^k08819/+; Cap-H2^z3-0019/+ flies in the iso82 genetic background were labeled with FISH probes to chromosomal positions 34D and 86C. Nurse cells maintain a pseudo-polytene structure, which is made evident by the DAPI-stained chromosomes and the clustered FISH spots indicating paired chromatids. Bars, 10 µm. (B and C) Two different Slimb alleles, UU11(B) and 3A1(C), carried on the same isogenic chromosome, were crossed with Smc4^k08819/+; Cap-H2^z3-0019/+, and triple heterozygous nurse cells were labeled with DAPI and FISH probes. (D) Stage 10 nurse cells triple labeled with FISH probes to chromosomal positions 34D, 86C, and 89D were quantified for number of spots per nucleus to determine the degree of polytene pairing. The number of FISH spots for each chromosomal position in Smc4^k08819/+; Cap-H2^z3-0019/+ was compared with Smc4^k08819/+; Cap-H2^z3-0019/+, Slimb^UU11/+ (*, P < 10⁻⁴; n = 50 nuclei) and Smc4^k08819/+; Cap-H2^z3-0019/+, Slimb^3A1/+ (**, P < 10⁻⁹; n = 50 nuclei). Error bars indicate SEM.