- Email: [email protected]
A New Histone at the Centromere?
Whereas the reflex principles described here are prototypical and well understood, it is quite likely that additional, as yet undiscovered neurophysiological reflexes will be mapped. The findings of Arima and colleagues represent a significant advance in this field and are likely to expand interest in new mechanisms and therapeutic strategies to specifically modulate the activity of neural circuits.
ACKNOWLEDGMENTS Kevin Tracey is a cofounder of Setpoint Medical.
Koopman, F.A., Stoof, S.P., Straub, R.H., Van Maanen, M.A., Vervoordeldonk, M.J., and Tak, P.P. (2011). Mol. Med. 17, 937–948. Nathan, C., and Ding, A. (2010). Cell 140, 871–882.
REFERENCES Arima, Y., Harada, M., Kamimura, D., Park, J.-H., Kawano, F., Yull, F.E., Kawamoto, T., Iwakura, Y., Betz, U.A., Ma´rquez, G., et al. (2012). Cell 148, this issue, 447–457.
Rosas-Ballina, M., Olofsson, P.S., Ochani, M., Valde´s-Ferrer, S.I., Levine, Y.A., Reardon, C., Tusche, M.W., Pavlov, V.A., Andersson, U., Chavan, S., et al. (2011). Science 334, 98–101. Tracey, K.J. (2009). Nat. Rev. Immunol. 9, 418–428.
A New Histone at the Centromere? Daniel R. Foltz1 and P. Todd Stukenberg1,* 1Department of Biochemistry and Molecular Genetics, University of Virginia, School of Medicine, Charlottesville, VA 22908, USA *Correspondence: [email protected] DOI 10.1016/j.cell.2012.01.023
The centromere is a classic system to study epigenetic specification, and most research has focused on a specialized histone variant, CENP-A, that is required for kinetochore assembly. Now Nishino et al. reveal a new level of complexity for centromeric chromatin, by showing that the kinetochore complex CENP-T-W-S-X shares structural and functional properties with canonical histones. The centromere is a unique domain of chromatin that is present as a single, stable locus on each chromosome. The best known function for centromeres is to assemble the mitotic kinetochore, a protein complex required for proper segregation of chromosomes. A central feature of all eukaryotic centromeres is a specialized nucleosome containing the CENP-A protein. Centromeres contain 16 other proteins, called the constitutive centromere-associated network (CCAN), which are recruited to the centromere by CENP-A nucleosomes. Until now, little structural data were available to explain how the constitutive centromere functions. In this issue, Nishino et al. (2012) perform a detailed structural analysis of four CCAN proteins: CENP-T, CENP-W, CENP-S, and CENP-X (CENP-T-W-S-X). The authors show that CENP-T-W-S-X forms a heterotetrameric complex capable of binding and wrapping centromeric DNA. This observation has significant implications for how kinetochores are anchored to centromeric chromatin
and how centromeres are epigenetically propagated. The structures also provide an exciting look into the atomic structure of the constitutive centromere. Canonical nucleosomes contain four types of histones: H2A H2B, H3, and H4. Four CCAN proteins, CENP-T, CENP-W, CENP-S, and CENP-X, are predicted to have histone folds, suggesting that a new histone might form at the centromere. However, histone folds are also found in non-nucleosomal proteins, such as the transcription factor proteins NFYB/C, NC2a-b, and components of the TAF complex. In these non-nucleosomal contexts, the histone folds can act as heterologous dimerization domains, as is the case for the TAF proteins, and also as DNA recognition motifs, as is the case for the NF-YB/C proteins. In contrast, the data presented by Nishino et al. suggest that the CENP-T-W-S-X complex shares both structure and functional properties of a canonical histone, suggesting that CENP-T-W-S-X could form a new type of nucleosome at centro-
394 Cell 148, February 3, 2012 ª2012 Elsevier Inc.
meres. To highlight this exciting hypothesis, we will review the similarities and differences of CENP-T-W-S-X and a canonical nucleosome. In both histones and non-DNA-binding proteins, two histone folds associate with each other to form a heterologous dimer. In histones, two of these dimers then come together to form tetramers (Figure 1A). Nishino et al. confirm the formation of the CENP-T-W and CENPS-X subcomplexes by in vitro reconstitution (Amano et al., 2009; Hori et al., 2008). They then solve the crystal structures of each subcomplex. These structures demonstrate that CENP-T-W and CENPS-X dimerize through their histone-fold domains. They combine the two subcomplexes to form the T-W-S-X heterotetramer, which contains a single copy of each protein. The interface between the CENP-T-W and -S-X heterodimers occurs in regions of CENP-T and CENP-S, which are similar to that observed in histone heterotetramers. An important difference between these complexes is that the
Figure 1. Structure and Cellular Roles of CENP-T-W-S-X Chromatin (A) Octameric nucleosomes are formed from the histone H3-H4 heterotetramer and H2A-H2B dimer intermediates. An octameric nucleosome protects 146 bp of DNA from degradation by micrococcal nuclease. Subnucleosomal complexes containing a heterotetramer of histones H3 and H4 can also form in vitro, and these complexes protect only 80 bp. CENP-S-X forms a heterotetramer, and CENP-T-W forms a heterodimer. When combined together, CENP-S-X-T-W forms a heterotetramer that contains a single copy of each protein. The CENP-T-W-S-X complex can wrap 100 bp. Here the CENP-S-X-T-W heterotetramer is illustrated as interacting across the H3-H4 heterodimer axis. However, the data by Nishino et al. (2012) do not preclude that the CENP-T-W and -S-X dimerization interface is analogous to that between H3-H4 and H2A-H2B heterodimers. (B) The CENP-T-W-S-X complex links centromeric chromatin to the mitotic kinetochore. The histone folds of the CENP-T-W-S-X complex are associated with DNA, whereas the CENP-T amino terminus is associated with the microtubule-binding complex NDC80. The yellow box depicts the remaining constitutive centromere-associated network (CCAN) complex, for which there are no structural data.
histone H3-H4 heterotetramer is a dimer of heterodimers and is therefore symmetric across the tetramerization interface. In contrast, the CENP-T-W-SX complex is asymmetric. Nishino et al. then demonstrate that tetramerization of CENP-T-W-S-X is important for its biological function; when they mutate the tetramerization interface in either CENP-T or CENP-S, recruitment to centromeres is lost in cells. An important feature of histones is their ability to wrap and organize DNA. The CENP-T-W-S-X complex, as well as the T-W and S-X subcomplexes, can similarly bind and bend DNA. The histone-fold motifs of nucleosomes organize the path of the DNA around the histone core through a series of protein-DNA interactions. Based on similarities of the CENPT-W complex to the canonical histones,
Nishino and colleagues identify residues that putatively bind and wrap DNA around the outside of the CENP-T-W-S-X tetramer. Mutating DNA-binding residues in the CENP-T-W half of the tetramer decreases centromere localization, suggesting that this interaction is required for centromere function. However, mutating the analogous DNA-binding residues in CENP-S-X does not alter DNA binding, leaving the path of the DNA around the CENP-T-W-S-X complex still somewhat unclear. Canonical histone proteins form nucleosomes capable of wrapping 146 bp of DNA around an octameric core, and nucleosome incorporation into plasmids introduces negative supercoiling. CENPT-W-S-X protects a smaller 100 bp region of DNA, and it can introduce supercoils, although it appears to wrap DNA less
efficiently than canonical H3-H4 heterotetramers. Given its protein content, CENP-T-W-S-X may be analogous to the subnucleosomal H3-H4-containing tetrasome, which wraps 80 bp of DNA (Lavelle and Prunell, 2007) (Figure 1A). It will be important to determine whether the CENP-T-W-S-X complex can form an octamer in the presence of DNA. The pulling forces that move chromosomes during mitosis require the kinetochore to tightly adhere to the underlying chromatin. Centromeres contain both CENP-A and canonical H3 nucleosomes. So the discovery that the CENP-T-WS-X complex is able bind DNA in a subnucleosome-like structure suggests that the pulling forces placed on the centromere may be distributed across three different chromatin-binding complexes: canonical H3-H4 nucleosomes, CENP-A-containing
Cell 148, February 3, 2012 ª2012 Elsevier Inc. 395
nucleosomes, and CENP-T-W-S-X nucleosomes (Figure 1B). The Ndc80 complex is the kinetochore’s direct attachment point to microtubules, and interestingly the forces put on Ndc80 are spread onto two different DNA-wrapping complexes. CENP-T is a large protein, and a portion of CENP-T, which is N-terminal to the histone domain, extends toward the kinetochore to directly bind the Ndc80 complex (Gascoigne et al., 2011). Another connection to the Ndc80 complex is built through CENP-C, which links the CENP-A nucleosome to Ndc80 through the Mis12 complex (Screpanti et al., 2011). The centromere must organize the kinetochore at a unique site on the chromosome. The location is not strictly determined by the underlying DNA sequence. One reason that CENP-A is thought to mediate this epigenetic inheritance is because CENP-A nucleosomes are highly stable and do not turn over (Jansen et al., 2007). It is still unknown whether the CENP-T-W-S-X complex is also involved in maintaining centromere identity. However, CENP-T is dependent on CENP-C for recruitment and is completely turned over each time the cell goes through the cell cycle (Prendergast et al., 2011). Given that CENP-T is not stably inherited through each round of cell division, the stable CENP-A nucleosome is still the best candidate to act as the epigenetic mark of the centromere. Recently, the CENP-S-X complex (which is also known as MHF1 and
MHF2) was also found in association with the Fanconi Anemia complex protein FANCM (Singh et al., 2010; Yan et al., 2010). FANCM responds to interstrand DNA crosslinks that prohibit replication fork progression. MHF1/2 binds doublestranded DNA and is required for the activity of the FANCM complex. Interestingly, two CENP-S-X complexes can dimerize and form multimers on DNA in vitro, although CENP-S-X prefers to form a heterotetramer with CENP-T-W. Nishino and colleagues suggest that CENP-S-X tetramers may mediate the DNA-damage response, whereas the CENP-S-X-T-W tetramer may specify the centromere. Although CENP-T and CENP-W were not copurified with FANCM, CENP-T has been localized to sites of DNA double-stranded breaks along with CENP-A (Zeitlin et al., 2009). An important area of future studies will be to define the commonalities in the character of chromatin at the centromere and at double-stranded breaks. The crystal structures of the CENP-TW-S-X complex provide important insight into the chromatin architecture that underlies centromere function and some DNA-damage responses. It is exciting that this complex shares many aspects of a canonical histone. However, a structural analysis of the DNA-bound CENP-T-W-S-X complex is required to ultimately determine whether it forms a second centromere-specific nucleosome structure.
396 Cell 148, February 3, 2012 ª2012 Elsevier Inc.
REFERENCES Amano, M., Suzuki, A., Hori, T., Backer, C., Okawa, K., Cheeseman, I.M., and Fukagawa, T. (2009). J. Cell Biol. 186, 173–182. Gascoigne, K.E., Takeuchi, K., Suzuki, A., Hori, T., Fukagawa, T., and Cheeseman, I.M. (2011). Cell 145, 410–422. Hori, T., Amano, M., Suzuki, A., Backer, C.B., Welburn, J.P., Dong, Y., McEwen, B.F., Shang, W.H., Suzuki, E., Okawa, K., et al. (2008). Cell 135, 1039–1052. Jansen, L.E., Black, B.E., Foltz, D.R., and Cleveland, D.W. (2007). J. Cell Biol. 176, 795–805. Lavelle, C., and Prunell, A. (2007). Cell Cycle 6, 2113–2119. Nishino, T., Takeuchi, K., Gascoigne, K.E., Suzuki, A., Hori, T., Oyama, T., Morikawa, K., Cheeseman, I.M., and Fukagawa, T. (2012). Cell 148, this issue, 487–501. Prendergast, L., van Vuuren, C., Kaczmarczyk, A., Doering, V., Hellwig, D., Quinn, N., Hoischen, C., Diekmann, S., and Sullivan, K.F. (2011). PLoS Biol. 9, e1001082. Screpanti, E., De Antoni, A., Alushin, G.M., Petrovic, A., Melis, T., Nogales, E., and Musacchio, A. (2011). Curr. Biol. 21, 391–398. Singh, T.R., Saro, D., Ali, A.M., Zheng, X.F., Du, C.H., Killen, M.W., Sachpatzidis, A., Wahengbam, K., Pierce, A.J., Xiong, Y., et al. (2010). Mol. Cell 37, 879–886. Yan, Z., Delannoy, M., Ling, C., Daee, D., Osman, F., Muniandy, P.A., Shen, X., Oostra, A.B., Du, H., Steltenpool, J., et al. (2010). Mol. Cell 37, 865–878. Zeitlin, S.G., Baker, N.M., Chapados, B.R., Soutoglou, E., Wang, J.Y., Berns, M.W., and Cleveland, D.W. (2009). Proc. Natl. Acad. Sci. USA 106, 15762–15767.
ACKNOWLEDGMENTS Kevin Tracey is a cofounder of Setpoint Medical.
Koopman, F.A., Stoof, S.P., Straub, R.H., Van Maanen, M.A., Vervoordeldonk, M.J., and Tak, P.P. (2011). Mol. Med. 17, 937–948. Nathan, C., and Ding, A. (2010). Cell 140, 871–882.
REFERENCES Arima, Y., Harada, M., Kamimura, D., Park, J.-H., Kawano, F., Yull, F.E., Kawamoto, T., Iwakura, Y., Betz, U.A., Ma´rquez, G., et al. (2012). Cell 148, this issue, 447–457.
Rosas-Ballina, M., Olofsson, P.S., Ochani, M., Valde´s-Ferrer, S.I., Levine, Y.A., Reardon, C., Tusche, M.W., Pavlov, V.A., Andersson, U., Chavan, S., et al. (2011). Science 334, 98–101. Tracey, K.J. (2009). Nat. Rev. Immunol. 9, 418–428.
A New Histone at the Centromere? Daniel R. Foltz1 and P. Todd Stukenberg1,* 1Department of Biochemistry and Molecular Genetics, University of Virginia, School of Medicine, Charlottesville, VA 22908, USA *Correspondence: [email protected] DOI 10.1016/j.cell.2012.01.023
The centromere is a classic system to study epigenetic specification, and most research has focused on a specialized histone variant, CENP-A, that is required for kinetochore assembly. Now Nishino et al. reveal a new level of complexity for centromeric chromatin, by showing that the kinetochore complex CENP-T-W-S-X shares structural and functional properties with canonical histones. The centromere is a unique domain of chromatin that is present as a single, stable locus on each chromosome. The best known function for centromeres is to assemble the mitotic kinetochore, a protein complex required for proper segregation of chromosomes. A central feature of all eukaryotic centromeres is a specialized nucleosome containing the CENP-A protein. Centromeres contain 16 other proteins, called the constitutive centromere-associated network (CCAN), which are recruited to the centromere by CENP-A nucleosomes. Until now, little structural data were available to explain how the constitutive centromere functions. In this issue, Nishino et al. (2012) perform a detailed structural analysis of four CCAN proteins: CENP-T, CENP-W, CENP-S, and CENP-X (CENP-T-W-S-X). The authors show that CENP-T-W-S-X forms a heterotetrameric complex capable of binding and wrapping centromeric DNA. This observation has significant implications for how kinetochores are anchored to centromeric chromatin
and how centromeres are epigenetically propagated. The structures also provide an exciting look into the atomic structure of the constitutive centromere. Canonical nucleosomes contain four types of histones: H2A H2B, H3, and H4. Four CCAN proteins, CENP-T, CENP-W, CENP-S, and CENP-X, are predicted to have histone folds, suggesting that a new histone might form at the centromere. However, histone folds are also found in non-nucleosomal proteins, such as the transcription factor proteins NFYB/C, NC2a-b, and components of the TAF complex. In these non-nucleosomal contexts, the histone folds can act as heterologous dimerization domains, as is the case for the TAF proteins, and also as DNA recognition motifs, as is the case for the NF-YB/C proteins. In contrast, the data presented by Nishino et al. suggest that the CENP-T-W-S-X complex shares both structure and functional properties of a canonical histone, suggesting that CENP-T-W-S-X could form a new type of nucleosome at centro-
394 Cell 148, February 3, 2012 ª2012 Elsevier Inc.
meres. To highlight this exciting hypothesis, we will review the similarities and differences of CENP-T-W-S-X and a canonical nucleosome. In both histones and non-DNA-binding proteins, two histone folds associate with each other to form a heterologous dimer. In histones, two of these dimers then come together to form tetramers (Figure 1A). Nishino et al. confirm the formation of the CENP-T-W and CENPS-X subcomplexes by in vitro reconstitution (Amano et al., 2009; Hori et al., 2008). They then solve the crystal structures of each subcomplex. These structures demonstrate that CENP-T-W and CENPS-X dimerize through their histone-fold domains. They combine the two subcomplexes to form the T-W-S-X heterotetramer, which contains a single copy of each protein. The interface between the CENP-T-W and -S-X heterodimers occurs in regions of CENP-T and CENP-S, which are similar to that observed in histone heterotetramers. An important difference between these complexes is that the
Figure 1. Structure and Cellular Roles of CENP-T-W-S-X Chromatin (A) Octameric nucleosomes are formed from the histone H3-H4 heterotetramer and H2A-H2B dimer intermediates. An octameric nucleosome protects 146 bp of DNA from degradation by micrococcal nuclease. Subnucleosomal complexes containing a heterotetramer of histones H3 and H4 can also form in vitro, and these complexes protect only 80 bp. CENP-S-X forms a heterotetramer, and CENP-T-W forms a heterodimer. When combined together, CENP-S-X-T-W forms a heterotetramer that contains a single copy of each protein. The CENP-T-W-S-X complex can wrap 100 bp. Here the CENP-S-X-T-W heterotetramer is illustrated as interacting across the H3-H4 heterodimer axis. However, the data by Nishino et al. (2012) do not preclude that the CENP-T-W and -S-X dimerization interface is analogous to that between H3-H4 and H2A-H2B heterodimers. (B) The CENP-T-W-S-X complex links centromeric chromatin to the mitotic kinetochore. The histone folds of the CENP-T-W-S-X complex are associated with DNA, whereas the CENP-T amino terminus is associated with the microtubule-binding complex NDC80. The yellow box depicts the remaining constitutive centromere-associated network (CCAN) complex, for which there are no structural data.
histone H3-H4 heterotetramer is a dimer of heterodimers and is therefore symmetric across the tetramerization interface. In contrast, the CENP-T-W-SX complex is asymmetric. Nishino et al. then demonstrate that tetramerization of CENP-T-W-S-X is important for its biological function; when they mutate the tetramerization interface in either CENP-T or CENP-S, recruitment to centromeres is lost in cells. An important feature of histones is their ability to wrap and organize DNA. The CENP-T-W-S-X complex, as well as the T-W and S-X subcomplexes, can similarly bind and bend DNA. The histone-fold motifs of nucleosomes organize the path of the DNA around the histone core through a series of protein-DNA interactions. Based on similarities of the CENPT-W complex to the canonical histones,
Nishino and colleagues identify residues that putatively bind and wrap DNA around the outside of the CENP-T-W-S-X tetramer. Mutating DNA-binding residues in the CENP-T-W half of the tetramer decreases centromere localization, suggesting that this interaction is required for centromere function. However, mutating the analogous DNA-binding residues in CENP-S-X does not alter DNA binding, leaving the path of the DNA around the CENP-T-W-S-X complex still somewhat unclear. Canonical histone proteins form nucleosomes capable of wrapping 146 bp of DNA around an octameric core, and nucleosome incorporation into plasmids introduces negative supercoiling. CENPT-W-S-X protects a smaller 100 bp region of DNA, and it can introduce supercoils, although it appears to wrap DNA less
efficiently than canonical H3-H4 heterotetramers. Given its protein content, CENP-T-W-S-X may be analogous to the subnucleosomal H3-H4-containing tetrasome, which wraps 80 bp of DNA (Lavelle and Prunell, 2007) (Figure 1A). It will be important to determine whether the CENP-T-W-S-X complex can form an octamer in the presence of DNA. The pulling forces that move chromosomes during mitosis require the kinetochore to tightly adhere to the underlying chromatin. Centromeres contain both CENP-A and canonical H3 nucleosomes. So the discovery that the CENP-T-WS-X complex is able bind DNA in a subnucleosome-like structure suggests that the pulling forces placed on the centromere may be distributed across three different chromatin-binding complexes: canonical H3-H4 nucleosomes, CENP-A-containing
Cell 148, February 3, 2012 ª2012 Elsevier Inc. 395
nucleosomes, and CENP-T-W-S-X nucleosomes (Figure 1B). The Ndc80 complex is the kinetochore’s direct attachment point to microtubules, and interestingly the forces put on Ndc80 are spread onto two different DNA-wrapping complexes. CENP-T is a large protein, and a portion of CENP-T, which is N-terminal to the histone domain, extends toward the kinetochore to directly bind the Ndc80 complex (Gascoigne et al., 2011). Another connection to the Ndc80 complex is built through CENP-C, which links the CENP-A nucleosome to Ndc80 through the Mis12 complex (Screpanti et al., 2011). The centromere must organize the kinetochore at a unique site on the chromosome. The location is not strictly determined by the underlying DNA sequence. One reason that CENP-A is thought to mediate this epigenetic inheritance is because CENP-A nucleosomes are highly stable and do not turn over (Jansen et al., 2007). It is still unknown whether the CENP-T-W-S-X complex is also involved in maintaining centromere identity. However, CENP-T is dependent on CENP-C for recruitment and is completely turned over each time the cell goes through the cell cycle (Prendergast et al., 2011). Given that CENP-T is not stably inherited through each round of cell division, the stable CENP-A nucleosome is still the best candidate to act as the epigenetic mark of the centromere. Recently, the CENP-S-X complex (which is also known as MHF1 and
MHF2) was also found in association with the Fanconi Anemia complex protein FANCM (Singh et al., 2010; Yan et al., 2010). FANCM responds to interstrand DNA crosslinks that prohibit replication fork progression. MHF1/2 binds doublestranded DNA and is required for the activity of the FANCM complex. Interestingly, two CENP-S-X complexes can dimerize and form multimers on DNA in vitro, although CENP-S-X prefers to form a heterotetramer with CENP-T-W. Nishino and colleagues suggest that CENP-S-X tetramers may mediate the DNA-damage response, whereas the CENP-S-X-T-W tetramer may specify the centromere. Although CENP-T and CENP-W were not copurified with FANCM, CENP-T has been localized to sites of DNA double-stranded breaks along with CENP-A (Zeitlin et al., 2009). An important area of future studies will be to define the commonalities in the character of chromatin at the centromere and at double-stranded breaks. The crystal structures of the CENP-TW-S-X complex provide important insight into the chromatin architecture that underlies centromere function and some DNA-damage responses. It is exciting that this complex shares many aspects of a canonical histone. However, a structural analysis of the DNA-bound CENP-T-W-S-X complex is required to ultimately determine whether it forms a second centromere-specific nucleosome structure.
396 Cell 148, February 3, 2012 ª2012 Elsevier Inc.
REFERENCES Amano, M., Suzuki, A., Hori, T., Backer, C., Okawa, K., Cheeseman, I.M., and Fukagawa, T. (2009). J. Cell Biol. 186, 173–182. Gascoigne, K.E., Takeuchi, K., Suzuki, A., Hori, T., Fukagawa, T., and Cheeseman, I.M. (2011). Cell 145, 410–422. Hori, T., Amano, M., Suzuki, A., Backer, C.B., Welburn, J.P., Dong, Y., McEwen, B.F., Shang, W.H., Suzuki, E., Okawa, K., et al. (2008). Cell 135, 1039–1052. Jansen, L.E., Black, B.E., Foltz, D.R., and Cleveland, D.W. (2007). J. Cell Biol. 176, 795–805. Lavelle, C., and Prunell, A. (2007). Cell Cycle 6, 2113–2119. Nishino, T., Takeuchi, K., Gascoigne, K.E., Suzuki, A., Hori, T., Oyama, T., Morikawa, K., Cheeseman, I.M., and Fukagawa, T. (2012). Cell 148, this issue, 487–501. Prendergast, L., van Vuuren, C., Kaczmarczyk, A., Doering, V., Hellwig, D., Quinn, N., Hoischen, C., Diekmann, S., and Sullivan, K.F. (2011). PLoS Biol. 9, e1001082. Screpanti, E., De Antoni, A., Alushin, G.M., Petrovic, A., Melis, T., Nogales, E., and Musacchio, A. (2011). Curr. Biol. 21, 391–398. Singh, T.R., Saro, D., Ali, A.M., Zheng, X.F., Du, C.H., Killen, M.W., Sachpatzidis, A., Wahengbam, K., Pierce, A.J., Xiong, Y., et al. (2010). Mol. Cell 37, 879–886. Yan, Z., Delannoy, M., Ling, C., Daee, D., Osman, F., Muniandy, P.A., Shen, X., Oostra, A.B., Du, H., Steltenpool, J., et al. (2010). Mol. Cell 37, 865–878. Zeitlin, S.G., Baker, N.M., Chapados, B.R., Soutoglou, E., Wang, J.Y., Berns, M.W., and Cleveland, D.W. (2009). Proc. Natl. Acad. Sci. USA 106, 15762–15767.