Condensin confers the longitudinal rigidity of chromosomes

Abstract

In addition to inter-chromatid cohesion, mitotic and meiotic chromatids must have three physical properties: compaction into 'threads' roughly co-linear with their DNA sequence, intra-chromatid cohesion determining their rigidity, and a mechanism to promote sister chromatid disentanglement. A fundamental issue in chromosome biology is whether a single molecular process accounts for all three features. There is universal agreement that a pair of Smc-kleisin complexes called condensin I and II facilitate sister chromatid disentanglement, but whether they also confer thread formation or longitudinal rigidity is either controversial or has never been directly addressed respectively. We show here that condensin II (beta-kleisin) has an essential role in all three processes during meiosis I in mouse oocytes and that its function overlaps with that of condensin I (gamma-kleisin), which is otherwise redundant. Pre-assembled meiotic bivalents unravel when condensin is inactivated by TEV cleavage, proving that it actually holds chromatin fibres together.

Figure 1

Condensin II but not condensin I is essential for meiosis. (a) Ncaph1^f/+ and Ncaph2^f/+ mice were crossed as indicated to address three different questions: the deletion efficiency of ZP3Cre on the respective flox allele, the fertility of Ncaph1^f/f Tg(ZP3Cre) and Ncaph2^f/f Tg(ZP3Cre) females, and the viability of the homozygous mice (Δ/Δ; 5 litters of 3 breeding pairs were analysed). In each case, three breeding pairs were analysed independently. (b) GVBD kinetics of wild-type, ΔNcaph1 and ΔNcaph2 oocytes. Oocytes were isolated in M16 supplemented with IBMX, after which IBXM was washed out and oocytes were followed by live cell confocal microscopy. The total number of oocytes are WT: n = 46, Ncaph1^f/f Tg(ZP3Cre): n = 45 and Ncaph2^f/f Tg(ZP3Cre): n = 42. Error bars represent mean ± s.d. (c) PBE efficiencies counted after 16 h of culture in vitro. The total number of oocytes are WT: n = 46, Ncaph1^f/f Tg(ZP3Cre): n = 45, Ncaph2^f/f Tg(ZP3Cre): n = 42, Ncaph1^f/f Tg(Gdf9–iCre): n = 35 and Ncaph2^f/f Tg(Gdf9–iCre): n = 112. Error bars represent mean ± s.d. (d) PBE kinetics of oocytes followed by live cell confocal microscopy. The total number of oocytes are WT: n = 46, Ncaph1^f/f Tg(ZP3Cre): n = 45 and Ncaph2^f/f Tg(ZP3Cre): n = 42. Error bars represent mean ± s.d. (e) The lack of PBE in Ncaph2^flox/flox Tg(Zp3Cre) oocytes and SAC activation. Oocytes were injected at the GV stage with RNA coding for securin–eGFP and wild-type CDC20 or CDC20R132A mutant. The degradation of securin–eGFP was monitored by low-resolution live cell imaging and quantified. Values from individual oocytes were normalized relative to that at GVBD (0 h), and mean and standard deviations of the population are plotted in arbitrary units, (a.u.), against time. The CDC20R132A mutant induces PBE with massive segregation defects in ΔNcaph2 oocytes (inset). The total number of oocytes are WT: n = 5, WT+CDC20: n = 6, WT+CDC20R132A: n = 6, Ncaph2^f/f Tg(ZP3Cre)+CDC20: n = 7. In b,d, oocytes were collected from one female of each genotype per experiment in three independent experiments. In c, oocytes were collected from one or two females of each genotype per experiment in three independent experiments. In e, oocytes were collected from one female of each genotype per experiment in two independent experiments.

Figure 2

Condensin II determines the morphology and rigidity of meiotic bivalents. Oocytes from Ncaph2^f/f females (wild-type Ncaph2) or Ncaph2^f/f Tg(ZP3Cre) females (ΔNcaph2) were injected at the GV stage with mRNA coding for H2B–mCherry to mark the whole chromosomes in red and TALE-mClover_MajSat to mark the pericentric repeats in green. Meiosis I progression was followed by live cell confocal imaging. Maximum intensity z projection images of the main time points are shown between 0.6 h post-GVBD and more than 10 h after, when the polar body (PB) is extruded by the wild-type control. The phenotype was observed in all oocytes analysed (oocyte number >50, female number >5, more than three experiments). Scale bar, 5 µm.

Figure 3

Deletion of Ncaph1 or Ncaph2 induces different condensation defects. (a,b) Chromosomes from oocytes isolated from Ncaph1^f/f Tg(ZP3Cre) (ΔNcaph1) (a) or Ncaph2^f/f Tg(ZP3Cre) (ΔNcaph2) (b) females were analysed by chromosome spreading 7 h (meiosis I) or 16 h (meiosis II) after GVBD. The depletion levels were analysed by immunofluorescence using antibodies directed against NCAPH1 or NCAPH2. The DNA was stained with DAPI. Oocytes from Ncaph1^f/f or Ncaph2^f/f females were used as the control (wild type). As 80% of ΔNcaph2 oocytes do not extrude the polar body, stretched bivalents are observed 16 h post-GVBD. The phenotypes were observed in all oocytes analysed (oocyte number >20, female number >5, more than three experiments). Scale bars, 5 µm.

Figure 4

Condensin is required for sister chromatid disentanglement even in the absence of cohesin. (a) Oocytes from Rec8^TEV/TEV Ncaph2^f/f (Rec8^TEV/TEV, WT Ncaph2) or Rec8^TEV/TEV, Ncaph2^f/f Tg(ZP3Cre) (Rec8^TEV/TEV, ΔNcaph2) females were injected at the GV stage with H2B–mCherry, securin–eGFP and Mad2 mRNA. Sixteen hours after GVBD, the oocytes arrested in metaphase I owing to MAD2 overexpression were injected with TEV protease mRNA and chromosomes were followed by live cell confocal imaging. Maximum-intensity z projection images of the indicated time points post TEV injection are shown. The phenotype was observed in all oocytes analysed (oocyte number = 15, female number 3, three experiments). (b) Oocytes from Rec8^TEV/TEV Ncaph2^f/f (Rec8^TEV/TEV, WT Ncaph2) or Rec8^TEV/TEV, Ncaph2^f/f Tg(ZP3Cre) (Rec8^TEV/TEV, ΔNcaph2) females were injected at the GV stage with mRNA coding for H2B–mCherry, eGFP–tubulin and TEV protease. Chromosomes were followed by live cell confocal imaging. Maximum-intensity z projection images of the indicated time points post-GVBD are shown. The phenotype was observed in all oocytes analysed (oocyte number = 19, female number = 3, three experiments). Scale bars, 10 µm.

Figure 5

Condensins I and II have overlapping functions. (a) PBE kinetics of oocytes isolated from females Ncaph1^f/f Ncaph2^f/f Tg(ZP3Cre). Oocytes from littermate Ncaph1^f/f Ncaph2^f/f females were used as the control. The two groups of oocytes were followed by live cell confocal microscopy to evaluate the time post-GVBD at which the polar body was extruded. Results are from two independent experiments realized using oocytes from one female of each genotype per experiment. The total number of oocytes are Ncaph1^f/f Ncaph2^f/f : n = 11 and Ncaph1^f/f Ncaph2^f/f Tg(ZP3Cre): n = 15. Error bars represent mean ± s.d. (b) Chromosome spreading 7 h post-GVBD of wild-type and ΔNcaph1 ΔNcaph2 oocytes stained with DAPI. The phenotype was observed in all oocytes analysed (oocyte number = 18, female number = 3, three experiments). Scale bar, 5 µm. (c,d). Oocytes from Ncaph1^f/f Ncaph2^f/f females (top) or Ncaph1^f/f Ncaph2^f/f Tg(ZP3Cre) females (bottom) were injected at the GV stage with mRNA coding for H2B–mCherry and TALE-mClover_MajSat. Meiosis I progression was followed by live cell confocal imaging. Maximum-intensity z projection images of the main time points are shown (PB, polar body). The phenotype was observed in all oocytes analysed (oocyte number = 32, female number = 3, three experiments). Scale bars, 7 µm.

Figure 6

Creation of a functional TEV-cleavable Ncaph2. (a) Oocytes were isolated from Ncaph2^f/f or Ncaph2^f/f Tg(ZP3Cre) females and injected at the GV stage with mRNA coding only for H2B–mCherry for wild type and ΔNcaph2. Two groups of ΔNcaph2 oocytes were injected with wild-type Ncaph2 or TEV-cleavable Ncaph2. The four groups of oocytes were followed by live cell confocal microscopy to evaluate the time post-GVBD at which the polar body was extruded. Results are from two independent experiments realized using oocytes from one female of each genotype per experiment. The total number of oocytes are wild type: n = 21, ΔNcaph2: n = 23, ΔNcaph2 + wild-type Ncaph2: n = 22, ΔNcaph2 + TEV-cleavable Ncaph2 n = 22. Error bars represent mean ± s.d. (b) ΔNcaph2 oocytes were injected at the GV stage with H2B–mCherry, H2B–mCherry + TEV-Ncaph2 or H2B–mCherry + TEV-Ncaph2 + TEV protease mRNA. Seven hours post-GVBD, chromosome spread analysis reveals that TEV-NCAPH2 is present on the chromosome scaffold (NCAPH2 immunostaining) and rescues the defects of chromosome organization (DAPI). However, no rescue is observed if the TEV protease-coding mRNA is co-injected. The phenotype was observed in all oocytes analysed (oocyte number = 10, female number = 2, two experiments). Scale bar, 5 µm.

Figure 7

Condensin maintains the morphology and rigidity of chromosomes. (a) Cartoon summarizing the experimental procedure. (b) Single Ncaph2 mutant (Ncaph2^f/f Tg(ZP3Cre), top panel) or double mutant oocytes (Ncaph1^f/f Ncaph2^f/f Tg(ZP3Cre), bottom panel) were injected at the GV stage with H2B–mCherry, TALE-mClover_MajSat, Mad2 and wild-type or TEV-cleavable NCAPH2. Sixteen hours after GVBD, the rescued oocytes arrested in metaphase I owing to MAD2 overexpression were injected with TEV protease mRNA and chromosomes were followed by live cell confocal imaging. Maximum-intensity z projection images of the indicated time points post TEV injection are shown. The phenotypes were observed in all oocytes analysed in three independent experiments: n> 15 for each category. Scale bar, 5 µm.

Figure 8

The stable association of condensin II with chromosomes maintains their structure. (a) Cartoon representing the experimental procedure. (b) Single Ncaph2 mutant oocytes were injected at the GV stage with H2B–mCherry, TALE-mClover_MajSat and wild-type or TEV-cleavable NCAPH2. Sixteen hours after GVBD, the rescued oocytes, naturally arrested in metaphase II were injected with TEV protease-coding mRNA and chromosomes were followed by live cell confocal imaging. Maximum-intensity z projection images of the indicated time points post TEV mRNA injection are shown. The phenotype was observed in all oocytes analysed (oocyte number >12, female number = 3, three experiments). Scale bar, 5 µm. (c) Chromosome spreads of the two different groups of oocytes, 2.5 h after TEV mRNA injection, showing the disruption of chromosome structure after TEV-NCAPH2 cleavage (top: rescue WT Ncaph2, bottom: rescue TEV-Ncaph2). The phenotype was observed in all oocytes analysed (oocyte number = 12, female number = 2, two experiments). Scale bar, 5 µm. (d) Slow recovery of condensin II after photobleaching. Ncaph2 oocytes were injected at the GV stage with H2B–mCherry-, MAD2- and NCAPH2–eGFP-coding mRNAs. After 16 h, half of the metaphase 1 chromosomes were photobleached (red circle) and the recovery of the eGFP signal was followed every hour. In the right panel, the average difference in mean eGFP fluorescence between bleached and non-bleached regions is plotted on a logarithmic scale (one experiment, n = 11 oocytes from two females). The trend line equation and the corresponding regression coefficient (R²) were obtained using Microsoft Excel software: y = y₀ * e^−kx. The half-life of the protein on the chromosome was calculated using the equation: T_1/2 = Ln(2)/k. Error bars represent s.d.