Files in this Data Supplement:
Fig. S1. Timeline of spindle bipolarisation and kinetochore reorientation during MI in MF-1 oocytes. (A) Schematic representation of criteria used for determining bipolarity. For bipolarity we reasoned that there should be two regions of microtubule-organising centres (MTOCs) that: (1) were diametrically opposed; (2) co-localised with opposite extremities of a bipolar-appearing spindle; and (3) were separated by an MTOC-free interval. (B) Confocal z-projections of oocytes immunostained for microtubules (β-tubulin) and MTOCs (γ-tubulin) during MI (n>15 per time point). Double arrowheads highlight MTOC-free regions at 6 and 8 hours post-GVBD. Shortly after GVBD, MTOCs adopted a spherical distribution. By 2 hours post-GVBD, γ-tubulin coalesced into discrete foci that appeared to be 'ejected' from the spindle, presaging the earliest change from the spherical spindle morphology. Based on β-tubulin staining, a bipolar-appearing spindle was apparent by 4-5 hours post-GVBD. Notably, however, at this stage γ-tubulin was not restricted to polar regions but was dispersed throughout the periphery and/or interior of the spindle in 88% of cases (n=24) and microtubules were not uniformly organised into anti-parallel running bundles. We considered such microtubule architecture a critical criterion of spindle bipolarity as chromosome movement between poles and equator would require uninterrupted tracks linking both aspects of the spindle. By 6 hours post-GVBD and beyond, MTOCs were predominantly restricted to diametrically opposed regions of the spindle (>95%; n=15), entirely consistent with recent data using pericentrin for labelling MTOCs (Breuer et al., 2010). In addition, by 6 hours post-GVBD, microtubules were now exclusively bundled into anti-parallel running longitudinal arrays. (C-E′) Confocal images of immunostained oocytes depicting kinetochore reorientation during spindle bipolarisation. (C,C′) At the microtubule ball stage the majority of bivalents are compact. (C′) A magnified image of a single z-section of the boxed region in C depicts a compact and a partially extended bivalent, with accompanying schematics below. A bivalent is schematised as comprising two homologous chromosomes (blue and green) united at a crossover (visible cytologically as the chiasma; grey arrow). Each homologue is itself composed of two sister chromatids united by cohesion (orange squares) and sister kinetochores (blue and green solid ovals) are constrained to act as a single unit (encircling blue and green ovals) during MI. Note that kinetochores of compact bivalents face in the same direction. (D) At 4 hours post-GVBD, when the spindle begins to take on a bipolar appearance, there is now a mixture of extended (white arrows) and compact (yellow arrows) bivalents. In the single z-section (Z1), it is clear that microtubule architecture is not yet exclusively composed of longitudinal microtubule bundles. Instead, there are 'windows' in the spindle that surround bivalents. In the case of the window highlighted by the dashed box, the enclosed bivalent is of a compact conformation. (E) By 6 hours post-GVBD, the spindle is now bipolar and comprises anti-parallel running longitudinal microtubule bundles with bivalents that are all extended and equatorially located. (E′) A magnified image of a single z-section of the boxed region in E depicts an extended bivalent with an accompanying schematic below. Note that kinetochores of extended bivalents face opposite directions in stark contrast with compact bivalents. Scale bars: 10 µm.
Fig. S2. Inter-kinetochore distances increase progressively during MI. (A-C) Immunostained oocytes illustrate the increase in inter-kinetochore distances from early (A) to late MI (B,C). Scale bar: 10 µm. (D) For measuring inter-kinetochore distances, oocytes were double labelled for ACA and DNA at 2, 4, 6 and 8 hours post-GVBD (see A-C). z-stacks were acquired at 1-2 µm intervals throughout the entire chromosome-containing region of the oocyte. The maximal distance (yellow lines, A-C) between the outer margins of ACA signals (white lines, A-C) was determined from pre-calibrated images acquired on the LSM 510 META by using the Line Drawing Mode in combination with the Measure function of the inbuilt Zeiss LSM software. For all stages, we only made measurements on bivalents whose two kinetochores could be seen in a single z-section and could be corroborated as being a homologous pair on the basis of decorating that bivalent�s extremities when the ACA and DNA channels were merged (see A-C). The mean inter-kinetochore distance was derived for oocytes at 2, 4, 6 and 8 hours post-GVBD, representing over 50 kinetochore pairs from 5-10 oocytes per time point, and assembled into a graph. Data are mean ± s.e.m.; *P<0.05 by Student's t-test. (A′-C′) Shown alongside immunostained images are schematic representations of chromosomal morphology and kinetochore orientation. Note that the increase in inter-kinetochore distances between 2 and 6 hours post-GVBD is largely the consequence of kinetochore reorientation (A′,B′), whereas the final increment between 6 and 8 hours post-GVBD after reorientation is complete reflects the distraction of kinetochore pairs towards opposite spindle poles (B′,C′).
Fig. S3. CENP-E undergoes net synthesis during MI and relocates from kinetochores to the spindle midzone at the metaphase I to anaphase I transition. (A) Samples (150 oocytes) were collected at 2, 4 and 8 hours post-GVBD as well as at 18 hours post-GVBD after oocytes have completed MI and are arrested at MII (Homer et al., 2009; Homer et al., 2005) and immunoblotted for CENP-E. Note that murine CENP-E is a large protein estimated to be 287 kDa (Weaver et al., 2003). GAPDH served as a loading control. Band intensities of CENP-E were quantified and normalised to values found at MII. (B) Oocytes at 2 and 8 hours post-GVBD were immunostained for DNA, CENP-E and Mad2. Note that both CENP-E and Mad2 localise to kinetochores at 2 hours post-GVBD but by 8 hours post-GVBD, when Mad2 is completely displaced from bivalents which have now become aligned at metaphase I, CENP-E remains localised to kinetochores. (C) Oocytes at 9 and 10 hours post-GVBD were immunostained for β-tubulin, DNA, CENP-E and Mad2. Note that CENP-E relocates to the spindle midzone and midbody at anaphase I and telophase I, respectively, and that Mad2 remains undetectable. Scale bars: 10 µm.
Fig. S4. Depletion of CENP-E using CENPEMO. (A) GV stage oocytes were microinjected with CENPEMO, maintained for 24 hours in 50 µM IBMX along with control uninjected oocytes before being washed into IBMX-free medium to allow resumption of MI. Samples (250 oocytes) were collected at 4 hours post-GVBD and immunoblotted for CENP-E. GAPDH served as a loading control. Band intensities of CENP-E were quantified and normalised to values found in control uninjected oocytes. (B,C) GV stage oocytes were microinjected with a standard control morpholino (ControlMO), maintained for 24 hours in 50 µM IBMX along with control uninjected oocytes before being washed into IBMX-free medium to allow resumption of MI. Oocytes were then fixed at 2 and 4 hours post-GVBD and immunostained for DNA and CENP-E. (D) GV stage oocytes were microinjected with CENPEMO, maintained for 24 hours in 50 µM IBMX before being washed into IBMX-free medium to allow resumption of MI. Oocytes were then fixed at 2 and 4 hours post-GVBD and immunostained for DNA, CENP-E and ACA. Note that from early in MI, CENP-E is largely undetectable in oocytes injected with CENPEMO (D, compare with B,C), whereas mock depletion had no discernible effects on CENP-E (C, compare with B). Scale bars: 10 µm.