- Email: [email protected]
C in Meiosis: Correct Me If I’m Right
Developmental Cell
Previews chromosome-specific topography that regulates gene expression chromosomewide. Previous studies have demonstrated that the DCC represses transcription by reducing RNA polymerase recruitment to promoters on X (Kruesi et al., 2013). How the modification of TAD boundaries could give rise to such a widespread effect remains to be seen. In other species, TAD boundaries are thought to define the genomic search space within which promoters are free to interact with enhancer or repressor elements. However, unlike mammals and Drosophila, few examples of long-range regulatory interactions have been reported in C. elegans, and a few kilobases of promoter sequence are often sufficient to recapitulate tissue-specific expression patterns in transgenic constructs (HuntNewbury et al., 2007). Positive supercoiling of DNA is one mechanism known to limit promoter accessibility to the transcriptional machinery, and it is interesting to note that condensins have been shown to promote supercoiling (Kimura and Hirano, 1997). Recently developed techniques that allow supercoiling to be assessed genome-wide (Naughton et al., 2013) could be used to determine whether the C. elegans X chromosome displays distinct, DCC-dependent supercoiling properties relative to autosomes.
This work substantially advances our understanding of worm dosage-compensation mechanisms and, more significantly, general principles of higherorder chromosome structure. The DCC is among the few trans-acting factors shown to influence TAD formation. Interestingly, the related cohesin complex, which is also homologous to condensin, is thought to partition fission yeast chromosomes into TAD-like structures (Mizuguchi et al., 2014). In mammalian cells, however, depletion of cohesin primarily affects interactions within TADs, rather than TAD boundaries (Seitan et al., 2013; Zuin et al., 2014). It is now pertinent to address whether condensin complexes function to partition topological domains in the interphase genomes of other organisms. More than 50 years after the first description of X inactivation by Mary Lyon, studies of dosage compensation continue to yield novel insights into fundamental mechanisms of gene regulation. The application of cutting-edge technology to further dissect these long-studied biological phenomena is likely to ensure that this trend continues apace. ACKNOWLEDGMENTS The authors thank Craig Nicol for assistance with graphic design.
REFERENCES Crane, E., Bian, Q., Patton McCord, R., Lajoie, B.R., Wheeler, B.S., Ralston, E.J., Uzawa, S., Dekker, J., and Meyer, B.J. (2015). Nature, in press, http://dx.doi.org/10.1038/nature14450. Published online June 1, 2015. Hirano, T. (2012). Genes Dev. 26, 1659–1678. Hunt-Newbury, R., Viveiros, R., Johnsen, R., Mah, A., Anastas, D., Fang, L., Halfnight, E., Lee, D., Lin, J., Lorch, A., et al. (2007). PLoS Biol. 5, e237. Kimura, K., and Hirano, T. (1997). Cell 90, 625–634. Kruesi, W.S., Core, L.J., Waters, C.T., Lis, J.T., and Meyer, B.J. (2013). eLife 2, e00808. Meyer, B.J. (2010). Curr. Opin. Genet. Dev. 20, 179–189. Mizuguchi, T., Fudenberg, G., Mehta, S., Belton, J.-M., Taneja, N., Folco, H.D., FitzGerald, P., Dekker, J., Mirny, L., Barrowman, J., and Grewal, S.I. (2014). Nature 516, 432–435. Naughton, C., Avlonitis, N., Corless, S., Prendergast, J.G., Mati, I.K., Eijk, P.P., Cockroft, S.L., Bradley, M., Ylstra, B., and Gilbert, N. (2013). Nat. Struct. Mol. Biol. 20, 387–395. Seitan, V.C., Faure, A.J., Zhan, Y., McCord, R.P., Lajoie, B.R., Ing-Simmons, E., Lenhard, B., Giorgetti, L., Heard, E., Fisher, A.G., et al. (2013). Genome Res. 23, 2066–2077. Zuin, J., Dixon, J.R., van der Reijden, M.I.J.A., Ye, Z., Kolovos, P., Brouwer, R.W.W., van de Corput, M.P.C., van de Werken, H.J.G., Knoch, T.A., van IJcken, W.F.J., et al. (2014). Cohesin and CTCF differentially affect chromatin architecture and gene expression in human cells. Proceedings of the National Academy of Sciences 111, 996–1001.
Aurora B/C in Meiosis: Correct Me If I’m Right Julien Dumont1,* 1Institut Jacques Monod, CNRS, UMR 7592, University Paris Diderot, Sorbonne Paris Cite ´ , 75205 Paris, France *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2015.05.018
In this issue of Developmental Cell, Yoshida et al. (2015) report that during meiosis I in mouse oocytes, the kinase Aurora B/C continuously destabilizes chromosome attachments to spindle microtubules, which potentially provides an explanation for the notably high error rate of chromosome segregation in mammalian oocytes. During cell division, chromosomes must properly attach to the spindle microtubules in order to be equally distributed between the newly forming cells. In mitosis, chromosomes are bioriented, with sister
kinetochores connected to microtubules emanating from opposite spindle poles. In contrast, during the first division of meiosis, sister kinetochores must establish monopolar attachments so that re-
combined homologous chromosomes are segregated (Dumont and Desai, 2012). A tight control of chromosomemicrotubule attachments is essential to avoid chromosome missegregation and
Developmental Cell 33, June 8, 2015 ª2015 Elsevier Inc. 499
Developmental Cell
Previews aneuploidy. Reporting in this issue of Developmental Cell, Yoshida et al. (2015) analyzed chromosome segregation in mouse oocytes to understand the origin of the particularly high rate of chromosome missegregation (20% of human oocytes are aneuploid) observed in mammalian female meiosis. During meiosis I (MI) in mouse oocytes, a slow maturation of chromosome-microtubule interactions occurs in two main phases: (1) chromosome stretching and (2) kinetochore-microtubule attachment stabilization (Kitajima et al., 2011). From 2 to 4 hr after nuclear envelope breakdown (NEBD), homologous chromosomes stretch along the spindle long axis. This stretched conformation is then stabilized over the last 3 hr of MI, up until anaphase onset. By analyzing cold-stable microtubules at these two stages, Yoshida et al. (2015) confirmed their previous observation that chromosome stretching occurs without a significant increase in the proportion of stable kinetochoremicrotubule attachments, which rises only during the stabilization phase (Kitajima et al., 2011). Aurora B kinase, the enzymatic subunit of the chromosomal passenger complex (CPC), is a known negative regulator of improper kinetochore-microtubule attachment during mitosis, when it functions by phosphorylating key kinetochore targets. In meiosis, an Aurora B-like kinase called Aurora C is expressed and participates in kinetochore-microtubule attachment error correction. Thus, the authors speculated that Aurora B/C activity might downregulate kinetochore-microtubule attachment stability during the stretching phase of MI. Two main models have been proposed for the Aurora B-dependent destabilization of improper attachments. The first and prevailing one is based on the spatial separation between Aurora B, which is concentrated at the inner centromere, and its kinetochore substrates. In this model, tension applied to sister kinetochores upon chromosome biorientation stabilizes attachment by spatially displacing kinetochore substrates from the negative regulation of the kinase at the inner centromere (Lampson and Cheeseman, 2011). The second model for attachment error correction relies on a tension-dependent intrinsic kinetochore structural transition that would allow for Aurora B kinet-
ochore target regulation independently of the kinase inner centromere localization (Cheerambathur and Desai, 2014). To test for spatial separation between Aurora B/C and kinetochore components and for a potential structural change such as intrakinetochore stretching during meiosis, Yoshida et al. (2015) compared the localization of INCENP (a regulatory subunit of the CPC), CENP-C (an inner kinetochore component), and Hec1 (an outer kinetochore component). While spatial separation and intrakinetochore stretching were evident in metaphase II-arrested oocytes with bioriented chromosomes, they were not detectable during MI even during the stabilization phase when stable attachments are set up. Instead, Aurora B/C kinase remained associated with the kinetochores throughout MI. These results show that, at least in MI, efficient kinetochore-microtubule attachments do not depend on spatial separation between Aurora B and its kinetochore targets. They also demonstrate that no major kinetochore structural change occurs during MI, which does not rule out a more subtle tension-dependent transition that would be detected by Aurora B/C. The spatial proximity between Aurora B/C and kinetochore components and the lack of detectable intrakinetochore stretching were intrinsic to MI chromosomes, as shown by the analysis of fused oocytes containing both MI and MII chromosomes aligned on the same spindle. Because the tension exerted on MI and MII kinetochores is identical in these fused oocytes, the presence of intrakinetochore stretching between MII, but not MI, kinetochores is consistent with MI centromeres being more stiff than their MII counterparts. The targeting of specialized meiotic cohesins to the core centromere of MI chromosomes, which promotes sister kinetochore co-orientation, could potentially explain this peculiar stiffness (Sakuno et al., 2009). Yoshida et al. (2015) also monitored the phosphorylation status of Knl1 and Hec1, two known Aurora B/C kinetochore targets. Consistent with the above results, the authors found that both proteins were continuously phosphorylated during MI. However, while their extent of phosphorylation increased continuously during the stretching phase, it dropped down to levels comparable to those in MII during the stabilization phase.
500 Developmental Cell 33, June 8, 2015 ª2015 Elsevier Inc.
To explain the apparent discrepancy between the phosphorylation status of Aurora B/C targets and the continuous presence of the kinase at the kinetochore during the stabilization phase, the authors searched for an activity that would oppose Aurora B/C activity and would progressively accumulate at the kinetochore. During mitosis, both criteria are fulfilled by the PP2A-B56 phosphatase complex, which also localizes at the kinetochore in mouse oocytes. By performing live-imaging experiments, Yoshida et al. (2015) demonstrated that the kinetochore level of PP2A-B56 progressively increased during MI. As in mitosis, this accumulation was at least in part downstream of the Cdk1- and Plk1dependent phosphorylation of the KARD domain of the BubR1 kinase. BubR1 is an essential component of the spindle assembly checkpoint (SAC), which verifies that all kinetochores are properly attached to the spindle microtubules before allowing the metaphase-toanaphase transition. In mitosis, BubR1 also has a SAC-independent function in the stabilization of kinetochore-microtubule attachments through the recruitment of PP2A at the kinetochore. Ectopic expression of phosphodeficient and phosphomimetic KARD domain mutant versions of BubR1 in mouse oocytes had the expected opposite effects on PP2A-B56 kinetochore recruitment and eventually on Knl1 phosphorylation (but surprisingly not on Hec1). Functionally, expression of the phosphomimetic and phosphodeficient mutants led to a respective increase and decrease of stable attachments. Together, these results provided an explanation for the role of Cdk1 as a timer that progressively stabilizes kinetochore-microtubule attachments in mouse oocytes (Davydenko et al., 2013). The progressive increase in Cdk1 activity would promote PP2A-B56 kinetochore recruitment through phosphorylation of the BubR1 KARD domain. The phosphatase would in turn oppose Aurora B/C activity and stabilize kinetochore-microtubule attachments late in MI. Although attractive, this simple model is probably not the full story. Indeed, a pool of PP2A-B56 that controls centromeric cohesion is recruited to the kinetochore downstream of the centromeric cohesion protector Sgo2 (Tanno et al., 2010). However, a role for this BubR1-independent
Developmental Cell
Previews pool of kinetochore PP2A-B56 in the control of kinetochore-microtubule attachments has not yet been tested. This pool could explain the relatively mild effects of the phosphomimetic and phosphodeficient BubR1 mutants on microtubule attachment, as well as their lack of effect on Hec1 phosphorylation. Moreover, a recent study demonstrated that defects of stable microtubule-kinetochore attachments in BubR1 knockout oocytes could be rescued by expression of a BubR1 mutant (BubR1 E406K) unable to localize at the kinetochore (Touati et al., 2015). Thus, BubR1 (and perhaps also PP2A-B56) has an additional role in kinetochore-microtubule attachment stabilization that does not require its kinetochore localization. It will be interesting to test whether a BubR1 phosphodeficient KARD mutant that cannot localize to the kinetochore is sufficient to rescue the chromosome attachment defects observed in the BubR1 knockout oocytes. Remarkably, Yoshida et al. (2015) showed that expression of the BubR1 phosphodeficient KARD mutant that delays stable attachments did not change the kinetics of chromosome stretching. This suggests that the stretching phase
occurs probably primarily through lateral interactions between kinetochores and microtubules. In mitosis, the kinetochore-localized minus-end-directed motor Dynein mediates these lateral interactions. Depletion of the Dynein kinetochore adaptor Spindly during meiosis should enable testing this hypothesis (Griffis et al., 2007). The authors also demonstrate that during the chromosomestretching phase, kinetochore-microtubule attachments are continuously destabilized, as shown by their dramatic increase following Aurora B/C inhibition. Strikingly, attachments that are formed upon Aurora B/C inhibition during this phase are correct and thus potentially able to sustain bivalent chromosome biorientation. In contrast, earlier Aurora B/C inhibition, before assembly of a bipolar spindle, leads to a dramatic increase of improper merotelic attachments. This result shows that, at least during the stretching phase in mouse oocytes, Aurora B/C continuously destabilizes potentially correct end-coupled kinetochore-microtubule attachments. Yoshida et al. (2015) propose that this seemingly counterproductive action of Aurora B/C during the chromosome-stretching phase
could explain the high rate of chromosome segregation errors observed in mammalian oocytes.
REFERENCES Cheerambathur, D.K., and Desai, A. (2014). Curr. Opin. Cell Biol. 26, 113–122. Davydenko, O., Schultz, R.M., and Lampson, M.A. (2013). J. Cell Biol. 202, 221–229. Dumont, J., and Desai, A. (2012). Trends Cell Biol. 22, 241–249. Griffis, E.R., Stuurman, N., and Vale, R.D. (2007). J. Cell Biol. 177, 1005–1015. Kitajima, T.S., Ohsugi, M., and Ellenberg, J. (2011). Cell 146, 568–581. Lampson, M.A., and Cheeseman, I.M. (2011). Trends Cell Biol. 21, 133–140. Sakuno, T., Tada, K., and Watanabe, Y. (2009). Nature 458, 852–858. Tanno, Y., Kitajima, T.S., Honda, T., Ando, Y., Ishiguro, K., and Watanabe, Y. (2010). Genes Dev. 24, 2169–2179. Touati, S.A., Buffin, E., Cladie`re, D., Hached, K., Rachez, C., van Deursen, J.M., and Wassmann, K. (2015). Nat. Commun. 6, 6946. Yoshida, S., Kaido, M., and Kitajima, T.S. (2015). Dev. Cell 33, this issue, 589–602.
Me´nage a Trois to Form the Tricellular Junction David Flores-Benitez1 and Elisabeth Knust1,* 1Max-Planck-Institute of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, 01307-Dresden, Germany *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2015.05.020
Tricellular junctions tightly seal epithelia at the corners of three cells. In this issue of Developmental Cell, Byri et al. (2015) show that Anakonda, a novel Drosophila transmembrane protein, contains an unusual tripartite extracellular domain organization, which explains the tripartite septum filling the tricellular junction, previously revealed by ultrastructure analysis. Metazoan life depends on the function of transporting epithelia, which selectively exchange substances with the environment. Tight junctions (TJs) in vertebrates and septate junctions (SJs) in arthropods limit the free diffusion of water and solutes through the space between epithelial cells, also known as the paracellular pathway. Early ultrastructural analyses
by freeze-fracture electron microscopy showed that TJs and SJs at bicellular junctions (BCJs) are composed of strands that run parallel to the epithelial plane and/or anastomose with each other (Claude and Goodenough, 1973; Staehelin, 1973). The strands are formed by rows of transmembrane proteins, which are tightly connected by their
extracellular domains. However, sites where three cells contact, called tricellular junctions (TCJs; Figure 1A), require particular proteins to seal the epithelium. In vertebrate TCJs, two vertical strands of central sealing elements spaced by 10 nm are associated tightly and laterally within individual plasma membranes, resulting in the formation of a central
Developmental Cell 33, June 8, 2015 ª2015 Elsevier Inc. 501
Previews chromosome-specific topography that regulates gene expression chromosomewide. Previous studies have demonstrated that the DCC represses transcription by reducing RNA polymerase recruitment to promoters on X (Kruesi et al., 2013). How the modification of TAD boundaries could give rise to such a widespread effect remains to be seen. In other species, TAD boundaries are thought to define the genomic search space within which promoters are free to interact with enhancer or repressor elements. However, unlike mammals and Drosophila, few examples of long-range regulatory interactions have been reported in C. elegans, and a few kilobases of promoter sequence are often sufficient to recapitulate tissue-specific expression patterns in transgenic constructs (HuntNewbury et al., 2007). Positive supercoiling of DNA is one mechanism known to limit promoter accessibility to the transcriptional machinery, and it is interesting to note that condensins have been shown to promote supercoiling (Kimura and Hirano, 1997). Recently developed techniques that allow supercoiling to be assessed genome-wide (Naughton et al., 2013) could be used to determine whether the C. elegans X chromosome displays distinct, DCC-dependent supercoiling properties relative to autosomes.
This work substantially advances our understanding of worm dosage-compensation mechanisms and, more significantly, general principles of higherorder chromosome structure. The DCC is among the few trans-acting factors shown to influence TAD formation. Interestingly, the related cohesin complex, which is also homologous to condensin, is thought to partition fission yeast chromosomes into TAD-like structures (Mizuguchi et al., 2014). In mammalian cells, however, depletion of cohesin primarily affects interactions within TADs, rather than TAD boundaries (Seitan et al., 2013; Zuin et al., 2014). It is now pertinent to address whether condensin complexes function to partition topological domains in the interphase genomes of other organisms. More than 50 years after the first description of X inactivation by Mary Lyon, studies of dosage compensation continue to yield novel insights into fundamental mechanisms of gene regulation. The application of cutting-edge technology to further dissect these long-studied biological phenomena is likely to ensure that this trend continues apace. ACKNOWLEDGMENTS The authors thank Craig Nicol for assistance with graphic design.
REFERENCES Crane, E., Bian, Q., Patton McCord, R., Lajoie, B.R., Wheeler, B.S., Ralston, E.J., Uzawa, S., Dekker, J., and Meyer, B.J. (2015). Nature, in press, http://dx.doi.org/10.1038/nature14450. Published online June 1, 2015. Hirano, T. (2012). Genes Dev. 26, 1659–1678. Hunt-Newbury, R., Viveiros, R., Johnsen, R., Mah, A., Anastas, D., Fang, L., Halfnight, E., Lee, D., Lin, J., Lorch, A., et al. (2007). PLoS Biol. 5, e237. Kimura, K., and Hirano, T. (1997). Cell 90, 625–634. Kruesi, W.S., Core, L.J., Waters, C.T., Lis, J.T., and Meyer, B.J. (2013). eLife 2, e00808. Meyer, B.J. (2010). Curr. Opin. Genet. Dev. 20, 179–189. Mizuguchi, T., Fudenberg, G., Mehta, S., Belton, J.-M., Taneja, N., Folco, H.D., FitzGerald, P., Dekker, J., Mirny, L., Barrowman, J., and Grewal, S.I. (2014). Nature 516, 432–435. Naughton, C., Avlonitis, N., Corless, S., Prendergast, J.G., Mati, I.K., Eijk, P.P., Cockroft, S.L., Bradley, M., Ylstra, B., and Gilbert, N. (2013). Nat. Struct. Mol. Biol. 20, 387–395. Seitan, V.C., Faure, A.J., Zhan, Y., McCord, R.P., Lajoie, B.R., Ing-Simmons, E., Lenhard, B., Giorgetti, L., Heard, E., Fisher, A.G., et al. (2013). Genome Res. 23, 2066–2077. Zuin, J., Dixon, J.R., van der Reijden, M.I.J.A., Ye, Z., Kolovos, P., Brouwer, R.W.W., van de Corput, M.P.C., van de Werken, H.J.G., Knoch, T.A., van IJcken, W.F.J., et al. (2014). Cohesin and CTCF differentially affect chromatin architecture and gene expression in human cells. Proceedings of the National Academy of Sciences 111, 996–1001.
Aurora B/C in Meiosis: Correct Me If I’m Right Julien Dumont1,* 1Institut Jacques Monod, CNRS, UMR 7592, University Paris Diderot, Sorbonne Paris Cite ´ , 75205 Paris, France *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2015.05.018
In this issue of Developmental Cell, Yoshida et al. (2015) report that during meiosis I in mouse oocytes, the kinase Aurora B/C continuously destabilizes chromosome attachments to spindle microtubules, which potentially provides an explanation for the notably high error rate of chromosome segregation in mammalian oocytes. During cell division, chromosomes must properly attach to the spindle microtubules in order to be equally distributed between the newly forming cells. In mitosis, chromosomes are bioriented, with sister
kinetochores connected to microtubules emanating from opposite spindle poles. In contrast, during the first division of meiosis, sister kinetochores must establish monopolar attachments so that re-
combined homologous chromosomes are segregated (Dumont and Desai, 2012). A tight control of chromosomemicrotubule attachments is essential to avoid chromosome missegregation and
Developmental Cell 33, June 8, 2015 ª2015 Elsevier Inc. 499
Developmental Cell
Previews aneuploidy. Reporting in this issue of Developmental Cell, Yoshida et al. (2015) analyzed chromosome segregation in mouse oocytes to understand the origin of the particularly high rate of chromosome missegregation (20% of human oocytes are aneuploid) observed in mammalian female meiosis. During meiosis I (MI) in mouse oocytes, a slow maturation of chromosome-microtubule interactions occurs in two main phases: (1) chromosome stretching and (2) kinetochore-microtubule attachment stabilization (Kitajima et al., 2011). From 2 to 4 hr after nuclear envelope breakdown (NEBD), homologous chromosomes stretch along the spindle long axis. This stretched conformation is then stabilized over the last 3 hr of MI, up until anaphase onset. By analyzing cold-stable microtubules at these two stages, Yoshida et al. (2015) confirmed their previous observation that chromosome stretching occurs without a significant increase in the proportion of stable kinetochoremicrotubule attachments, which rises only during the stabilization phase (Kitajima et al., 2011). Aurora B kinase, the enzymatic subunit of the chromosomal passenger complex (CPC), is a known negative regulator of improper kinetochore-microtubule attachment during mitosis, when it functions by phosphorylating key kinetochore targets. In meiosis, an Aurora B-like kinase called Aurora C is expressed and participates in kinetochore-microtubule attachment error correction. Thus, the authors speculated that Aurora B/C activity might downregulate kinetochore-microtubule attachment stability during the stretching phase of MI. Two main models have been proposed for the Aurora B-dependent destabilization of improper attachments. The first and prevailing one is based on the spatial separation between Aurora B, which is concentrated at the inner centromere, and its kinetochore substrates. In this model, tension applied to sister kinetochores upon chromosome biorientation stabilizes attachment by spatially displacing kinetochore substrates from the negative regulation of the kinase at the inner centromere (Lampson and Cheeseman, 2011). The second model for attachment error correction relies on a tension-dependent intrinsic kinetochore structural transition that would allow for Aurora B kinet-
ochore target regulation independently of the kinase inner centromere localization (Cheerambathur and Desai, 2014). To test for spatial separation between Aurora B/C and kinetochore components and for a potential structural change such as intrakinetochore stretching during meiosis, Yoshida et al. (2015) compared the localization of INCENP (a regulatory subunit of the CPC), CENP-C (an inner kinetochore component), and Hec1 (an outer kinetochore component). While spatial separation and intrakinetochore stretching were evident in metaphase II-arrested oocytes with bioriented chromosomes, they were not detectable during MI even during the stabilization phase when stable attachments are set up. Instead, Aurora B/C kinase remained associated with the kinetochores throughout MI. These results show that, at least in MI, efficient kinetochore-microtubule attachments do not depend on spatial separation between Aurora B and its kinetochore targets. They also demonstrate that no major kinetochore structural change occurs during MI, which does not rule out a more subtle tension-dependent transition that would be detected by Aurora B/C. The spatial proximity between Aurora B/C and kinetochore components and the lack of detectable intrakinetochore stretching were intrinsic to MI chromosomes, as shown by the analysis of fused oocytes containing both MI and MII chromosomes aligned on the same spindle. Because the tension exerted on MI and MII kinetochores is identical in these fused oocytes, the presence of intrakinetochore stretching between MII, but not MI, kinetochores is consistent with MI centromeres being more stiff than their MII counterparts. The targeting of specialized meiotic cohesins to the core centromere of MI chromosomes, which promotes sister kinetochore co-orientation, could potentially explain this peculiar stiffness (Sakuno et al., 2009). Yoshida et al. (2015) also monitored the phosphorylation status of Knl1 and Hec1, two known Aurora B/C kinetochore targets. Consistent with the above results, the authors found that both proteins were continuously phosphorylated during MI. However, while their extent of phosphorylation increased continuously during the stretching phase, it dropped down to levels comparable to those in MII during the stabilization phase.
500 Developmental Cell 33, June 8, 2015 ª2015 Elsevier Inc.
To explain the apparent discrepancy between the phosphorylation status of Aurora B/C targets and the continuous presence of the kinase at the kinetochore during the stabilization phase, the authors searched for an activity that would oppose Aurora B/C activity and would progressively accumulate at the kinetochore. During mitosis, both criteria are fulfilled by the PP2A-B56 phosphatase complex, which also localizes at the kinetochore in mouse oocytes. By performing live-imaging experiments, Yoshida et al. (2015) demonstrated that the kinetochore level of PP2A-B56 progressively increased during MI. As in mitosis, this accumulation was at least in part downstream of the Cdk1- and Plk1dependent phosphorylation of the KARD domain of the BubR1 kinase. BubR1 is an essential component of the spindle assembly checkpoint (SAC), which verifies that all kinetochores are properly attached to the spindle microtubules before allowing the metaphase-toanaphase transition. In mitosis, BubR1 also has a SAC-independent function in the stabilization of kinetochore-microtubule attachments through the recruitment of PP2A at the kinetochore. Ectopic expression of phosphodeficient and phosphomimetic KARD domain mutant versions of BubR1 in mouse oocytes had the expected opposite effects on PP2A-B56 kinetochore recruitment and eventually on Knl1 phosphorylation (but surprisingly not on Hec1). Functionally, expression of the phosphomimetic and phosphodeficient mutants led to a respective increase and decrease of stable attachments. Together, these results provided an explanation for the role of Cdk1 as a timer that progressively stabilizes kinetochore-microtubule attachments in mouse oocytes (Davydenko et al., 2013). The progressive increase in Cdk1 activity would promote PP2A-B56 kinetochore recruitment through phosphorylation of the BubR1 KARD domain. The phosphatase would in turn oppose Aurora B/C activity and stabilize kinetochore-microtubule attachments late in MI. Although attractive, this simple model is probably not the full story. Indeed, a pool of PP2A-B56 that controls centromeric cohesion is recruited to the kinetochore downstream of the centromeric cohesion protector Sgo2 (Tanno et al., 2010). However, a role for this BubR1-independent
Developmental Cell
Previews pool of kinetochore PP2A-B56 in the control of kinetochore-microtubule attachments has not yet been tested. This pool could explain the relatively mild effects of the phosphomimetic and phosphodeficient BubR1 mutants on microtubule attachment, as well as their lack of effect on Hec1 phosphorylation. Moreover, a recent study demonstrated that defects of stable microtubule-kinetochore attachments in BubR1 knockout oocytes could be rescued by expression of a BubR1 mutant (BubR1 E406K) unable to localize at the kinetochore (Touati et al., 2015). Thus, BubR1 (and perhaps also PP2A-B56) has an additional role in kinetochore-microtubule attachment stabilization that does not require its kinetochore localization. It will be interesting to test whether a BubR1 phosphodeficient KARD mutant that cannot localize to the kinetochore is sufficient to rescue the chromosome attachment defects observed in the BubR1 knockout oocytes. Remarkably, Yoshida et al. (2015) showed that expression of the BubR1 phosphodeficient KARD mutant that delays stable attachments did not change the kinetics of chromosome stretching. This suggests that the stretching phase
occurs probably primarily through lateral interactions between kinetochores and microtubules. In mitosis, the kinetochore-localized minus-end-directed motor Dynein mediates these lateral interactions. Depletion of the Dynein kinetochore adaptor Spindly during meiosis should enable testing this hypothesis (Griffis et al., 2007). The authors also demonstrate that during the chromosomestretching phase, kinetochore-microtubule attachments are continuously destabilized, as shown by their dramatic increase following Aurora B/C inhibition. Strikingly, attachments that are formed upon Aurora B/C inhibition during this phase are correct and thus potentially able to sustain bivalent chromosome biorientation. In contrast, earlier Aurora B/C inhibition, before assembly of a bipolar spindle, leads to a dramatic increase of improper merotelic attachments. This result shows that, at least during the stretching phase in mouse oocytes, Aurora B/C continuously destabilizes potentially correct end-coupled kinetochore-microtubule attachments. Yoshida et al. (2015) propose that this seemingly counterproductive action of Aurora B/C during the chromosome-stretching phase
could explain the high rate of chromosome segregation errors observed in mammalian oocytes.
REFERENCES Cheerambathur, D.K., and Desai, A. (2014). Curr. Opin. Cell Biol. 26, 113–122. Davydenko, O., Schultz, R.M., and Lampson, M.A. (2013). J. Cell Biol. 202, 221–229. Dumont, J., and Desai, A. (2012). Trends Cell Biol. 22, 241–249. Griffis, E.R., Stuurman, N., and Vale, R.D. (2007). J. Cell Biol. 177, 1005–1015. Kitajima, T.S., Ohsugi, M., and Ellenberg, J. (2011). Cell 146, 568–581. Lampson, M.A., and Cheeseman, I.M. (2011). Trends Cell Biol. 21, 133–140. Sakuno, T., Tada, K., and Watanabe, Y. (2009). Nature 458, 852–858. Tanno, Y., Kitajima, T.S., Honda, T., Ando, Y., Ishiguro, K., and Watanabe, Y. (2010). Genes Dev. 24, 2169–2179. Touati, S.A., Buffin, E., Cladie`re, D., Hached, K., Rachez, C., van Deursen, J.M., and Wassmann, K. (2015). Nat. Commun. 6, 6946. Yoshida, S., Kaido, M., and Kitajima, T.S. (2015). Dev. Cell 33, this issue, 589–602.
Me´nage a Trois to Form the Tricellular Junction David Flores-Benitez1 and Elisabeth Knust1,* 1Max-Planck-Institute of Molecular Cell Biology and Genetics, Pfotenhauerstrasse 108, 01307-Dresden, Germany *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2015.05.020
Tricellular junctions tightly seal epithelia at the corners of three cells. In this issue of Developmental Cell, Byri et al. (2015) show that Anakonda, a novel Drosophila transmembrane protein, contains an unusual tripartite extracellular domain organization, which explains the tripartite septum filling the tricellular junction, previously revealed by ultrastructure analysis. Metazoan life depends on the function of transporting epithelia, which selectively exchange substances with the environment. Tight junctions (TJs) in vertebrates and septate junctions (SJs) in arthropods limit the free diffusion of water and solutes through the space between epithelial cells, also known as the paracellular pathway. Early ultrastructural analyses
by freeze-fracture electron microscopy showed that TJs and SJs at bicellular junctions (BCJs) are composed of strands that run parallel to the epithelial plane and/or anastomose with each other (Claude and Goodenough, 1973; Staehelin, 1973). The strands are formed by rows of transmembrane proteins, which are tightly connected by their
extracellular domains. However, sites where three cells contact, called tricellular junctions (TCJs; Figure 1A), require particular proteins to seal the epithelium. In vertebrate TCJs, two vertical strands of central sealing elements spaced by 10 nm are associated tightly and laterally within individual plasma membranes, resulting in the formation of a central
Developmental Cell 33, June 8, 2015 ª2015 Elsevier Inc. 501