Cell-cycle-dependent Xenopus TRF1 recruitment to telomere chromatin regulated by Polo-like kinase
2006; Springer Nature; Volume: 25; Issue: 3 Linguagem: Inglês
10.1038/sj.emboj.7600964
ISSN1460-2075
AutoresAtsuya Nishiyama, Keiko Muraki, Motoki Saito, Keita Ohsumi, Takeo Kishimoto, Fuyuki Ishikawa,
Tópico(s)DNA Repair Mechanisms
ResumoArticle19 January 2006free access Cell-cycle-dependent Xenopus TRF1 recruitment to telomere chromatin regulated by Polo-like kinase Atsuya Nishiyama Atsuya Nishiyama Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Keiko Muraki Keiko Muraki Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Motoki Saito Motoki Saito Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Keita Ohsumi Keita Ohsumi Laboratory of Cell and Developmental Biology, Graduate School of Bioscience, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama, Japan Search for more papers by this author Takeo Kishimoto Takeo Kishimoto Laboratory of Cell and Developmental Biology, Graduate School of Bioscience, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama, Japan Search for more papers by this author Fuyuki Ishikawa Corresponding Author Fuyuki Ishikawa Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Atsuya Nishiyama Atsuya Nishiyama Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Keiko Muraki Keiko Muraki Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Motoki Saito Motoki Saito Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Keita Ohsumi Keita Ohsumi Laboratory of Cell and Developmental Biology, Graduate School of Bioscience, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama, Japan Search for more papers by this author Takeo Kishimoto Takeo Kishimoto Laboratory of Cell and Developmental Biology, Graduate School of Bioscience, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama, Japan Search for more papers by this author Fuyuki Ishikawa Corresponding Author Fuyuki Ishikawa Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan Search for more papers by this author Author Information Atsuya Nishiyama1, Keiko Muraki1, Motoki Saito1, Keita Ohsumi2, Takeo Kishimoto2 and Fuyuki Ishikawa 1 1Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa-Oiwake-cho, Sakyo-ku, Kyoto, Japan 2Laboratory of Cell and Developmental Biology, Graduate School of Bioscience, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama, Japan *Corresponding author. Laboratory of Cell Cycle Regulation, Department of Gene Mechanisms, Graduate School of Biostudies, Kyoto University, Kitashirakawa Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan. Tel.: +81 75 753 4195; Fax: +81 75 753 4197; E-mail: [email protected] The EMBO Journal (2006)25:575-584https://doi.org/10.1038/sj.emboj.7600964 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Telomeres are regulated by a homeostatic mechanism that includes telomerase and telomeric repeat binding proteins, TRF1 and TRF2. Recently, it has been hypothesized that telomeres assume distinct configurations in a cell-cycle-dependent manner, although direct biochemical evidence is lacking. Here we demonstrated that Xenopus TRF1 (xTRF1) associates with telomere chromatin specifically in mitotic Xenopus egg extracts, and dissociates from it upon mitotic exit. Both the N-terminal TRF-homology (TRFH) domain and the linker region connecting the TRFH domain and the C-terminal Myb domain are required for this cell-cycle-dependent association of xTRF1 with chromatin. In contrast, Xenopus TRF2 (xTRF2) associates with chromatin throughout the cell cycle. We showed that Polo-like kinase (Plx1) phosphorylates xTRF1 in vitro. Moreover, the mitotic xTRF1–chromatin association was significantly impaired when Plx1 was immunodepleted from the extracts. Finally, high telomerase activities were detected in association with replicating interphase chromatin compared with mitotic chromatin. These results indicate that telomere chromatin is actively regulated by cell-cycle-dependent processes, and provide an insight for understanding how telomeres undergo DNA metabolisms during the cell cycle. Introduction Eukaryotic chromosomes terminate in nucleoprotein complexes, termed telomeres, that are composed of arrays of telomeric repetitive sequences and telomere-binding proteins (reviewed in de Lange, 2005). One strand of telomeric DNA, the 3′-end of which faces the DNA end, is G-rich (TTAGGG repeats in vertebrates, G-strand), whereas the other strand is C-rich (C-strand). At the extreme end of telomeric DNA, the G-strand extends its 3′-terminus as a single-stranded DNA, a structure called the G-tail (Wellinger et al, 1993). Telomeres protect the ends of chromosomes from noxious reactions such as end-to-end fusion and degradation. Telomeric DNA, together with specific and nonspecific telomere-binding proteins, participates in forming higher ordered structures that protect chromosomal ends (reviewed in de Lange, 2005). Telomeric DNA becomes shortened every time cells divide, due to incomplete DNA replication at the most terminal lagging-strand synthesis (the end replication problem (Harley, 1991; Ohki et al, 2001)). Telomerase, a specialized form of reverse transcriptase containing its own template RNA, counteracts telomere shortening (reviewed in Cech, 2004; Chan and Blackburn, 2004). Two components, namely, the catalytic subunit, TERT (telomerase reverse transcriptase), and the RNA template, TR (telomerase RNA), are essential for telomerase activity measured in vitro. It is only in some specific types of normal human cells, including germ cells, actively dividing embryonal cells and adult progenitor cells of proliferating tissues, that telomerase activity has been detected. There is mounting evidence that telomerase activity is regulated in cis at individual telomeres by the number of associated telomeric-DNA-binding proteins (the protein-counting model) (Marcand et al, 1997). Such proteins include Rap1p in the budding yeast, Saccharomyces cerevisiae; Taz1 in the fission yeast, Schizosaccharomyces pombe; and TRF1 (TTAGGG repeat binding factor 1) in mammals. These proteins share the Myb-like domain, and Taz1 and TRF1 form homodimers that bind to telomeric DNA through the TRF-homology (TRFH) domain (Bianchi et al, 1997; Spink et al, 2000). TRF1 negatively regulates telomere length, as evidenced by observations that telomere lengths are decreased and increased by the overexpression of wild-type and dominant-negative mutants, respectively (van Steensel and de Lange, 1997). TRF2, a paralog of TRF1, also binds to telomeric DNA as a homodimer through the Myb-like domain (Bilaud et al, 1997; Broccoli et al, 1997). TRF2 plays an important role in protecting chromosomal ends, rather than controlling telomerase reaction, because the dominant-negative mutant expression induces chromosomal end-to-end fusion (van Steensel et al, 1998). It has been proposed that these telomere-binding proteins regulate the accessibility of telomerase as well as protect chromosomal ends by determining the configuration of the telomere complex. Specifically, it has been proposed that TRF1 and TRF2 collaborate to form a terminal loop structure called the t-loop, thereby concealing the 3′-end of the G-tail from the action of telomerase or other enzymatic activities (Griffith et al, 1999; de Lange, 2004). One recent study of budding yeast has demonstrated that short telomeres have a higher probability of being extended by telomerase than long telomeres, suggesting the presence of two telomere configurations, one accessible and the other inaccessible to telomerase (Teixeira et al, 2004). In addition, another study has provided evidence implicating the presence of an intermediate state of telomere configuration that contains TRF2 but lacks TRF1, thereby protecting the chromosomal ends, yet allowing access by telomerase (Chang et al, 2003). However, it is not known how and when in the cell cycle these distinct telomere configurations transit among each other. As the G-strand synthesis by telomerase is intimately coupled with the C-strand synthesis by the conventional lagging-strand DNA synthesis (Nakamura et al, 2005; reviewed in Price, 1997), the telomerase reaction should proceed coordinately with DNA replication in S phase. Indeed, budding yeast telomerase elongates telomeres in late S phase (Marcand et al, 2000). This argument suggests that telomere configuration is regulated in a cell-cycle-dependent manner, such that it becomes accessible to telomerase only transiently within the narrow window of S phase when telomere replication occurs. Here we investigated whether telomere chromatin is dynamically regulated during the cell cycle through biochemical approaches using unfertilized Xenopus egg extracts, in which high levels of telomerase and cyclin B–Cdc2 kinase activities are retained and cyclin B destruction is triggered by the addition of Ca2+ (Murray, 1991; Mantell and Greider, 1994). We found that Xenopus TRF1 (xTRF1) dynamically associates with mitotic telomere chromatin and actively dissociates from interphase telomere chromatin in Xenopus egg extracts. Moreover, we demonstrated that Polo-like kinase (Plx1), a Xenopus member of the Plx1 family, plays a critical role in regulating the dynamic behavior of TRF1. We also provide evidence that Xenopus TRF2 (xTRF2) is bound to telomere chromatin throughout the cell cycle and phosphorylated in a cell-cycle-dependent manner. These results provide a new avenue for understanding the regulatory mechanism of telomere configurations. Results Identification of xTRF1 and xTRF2 We searched the EST database for the Xenopus homologs of TRF1 and TRF2 (xTRF1 and xTRF2) and obtained Xenopus EST sequences having high homology with human TRF1 (hTRF1) or TRF2 (hTRF2) (accession nos. BG552350 and BJ638816, respectively). Several independent clones containing the EST sequence were isolated, and the composite full-length xTRF1 and xTRF2 cDNA sequences encoding open reading frames of 420 and 468 amino acids, respectively, were deduced (Supplementary Figure 1). We noticed that whereas mouse and human TRF1 and TRF2 proteins possess amino-terminal regions that are rich in acidic and basic amino acids, respectively, xTRF1 and xTRF2 lack such regions. We demonstrated that recombinant xTRF1 and xTRF2 produced in reticulocyte lysates or Escherichia coli form homodimers (data not shown) and bind to double-stranded telomeric DNA in electrophoretic mobility shift assay (EMSA) as human TRF proteins do (Supplementary Figure 2). Chromatin association of xTRF1 and xTRF2 during the cell cycle in Xenopus egg extracts When Xenopus sperm (XSP) chromatin is incubated with extracts prepared from unfertilized Xenopus eggs arrested at meiotic metaphase II (CSF extracts for cytostatic-factor-arrested extracts), condensed mitotic chromatin is assembled (Murray, 1991). Therefore, CSF extracts recapitulate M phase events. The addition of Ca2+ to CSF extracts induces cyclin B destruction through the ubiquitin–APC–proteasome pathway, leading to the inactivation of cyclin B–Cdc2 kinase and a transition to interphase (interphase extracts) (Glotzer et al, 1991). Accordingly, upon the addition of Ca2+, the condensed chromatin in CSF extracts becomes decondensed, forms the nuclear envelope, and undergoes DNA replication. Using this system, we investigated how the chromosomal association of Xenopus TRF proteins is regulated during the cell cycle. Despite repeated trials, we were unable to prepare anti-xTRF1 antibodies that are sufficiently sensitive to detect endogenous xTRF1. In contrast, endogenous xTRF2 was successfully detected by specific antibodies (see below). To facilitate the detection of xTRF1 in egg extracts, 35S-xTRF1 was in vitro translated in rabbit reticulocyte lysate (RRL) and added together with sperm chromatin to CSF or interphase extracts, and incubation was carried out for 1 h. Chromatin was isolated from the extracts by centrifugation, and chromatin-bound 35S-xTRF1 and endogenous xTRF2 were analyzed by autoradiography and immunoblotting, respectively. Similar amounts of histone H2B were recovered from both chromatins in CSF and interphase extracts, which served as an internal control of protein recovery (Figure 1A, bottom panel, lanes 3 and 4). We found that xTRF1 was specifically recovered with mitotic chromatin in CSF extracts, but not with interphase chromatin, as revealed by autoradiography (Figure 1A, top panel, lanes 3 and 4). Immunoblot analysis of endogenous xTRF2, in contrast, showed association with chromatin in both mitotic and interphase extracts, indicating that xTRF2 associates with chromatin throughout the cell cycle (Figure 1A, middle panel, lanes 3 and 4). We noticed that xTRF2 derived from mitotic chromatin migrated more slowly than that derived from interphase chromatin on SDS–PAGE (Figure 1A, middle panel, lanes 3 and 4). This is due, at least in part, to the phosphorylation of xTRF2, as will be described below. Figure 1.Chromatin binding of Xenopus TRF proteins during the cell cycle. (A) Chromatin binding of xTRF1 and xTRF2 in mitotic or interphase extracts. XSP chromatin was incubated with CSF (M, lanes 1 and 3) or interphase (I, lanes 2 and 4) extracts containing 35S-xTRF1 for 1 h. Chromatin was isolated by centrifugation on a sucrose cushion and the chromatin-bound proteins were analyzed for xTRF1 by phosphorimaging (top panel), and for xTRF2 and H2B by immunoblotting (middle and bottom panels). (B) Addition of nondegradable cyclin B to interphase extracts induces xTRF1 association with chromatin. Sperm chromatin was incubated with CSF (M, lanes 2–5) or interphase (I, lanes 6–13) extracts containing 35S-xTRF1. After 2 h incubation, the interphase extracts were supplemented with recombinant ΔN-cyclin B (ΔN-cycB) and incubated for an additional 2 h (I+ΔN-cycB, lanes 10–13). CSF extracts incubated with the labeled xTRF1 but not with sperm chromatin for 2 h were analyzed in lane 1. Aliquots were taken at the indicated time points during the incubation and chromatin-bound proteins were analyzed for xTRF1 by phosphorimaging (top panel), for histone H2A by immunoblotting (middle panel; the upper band is H2A.X, whereas the lower band is H2A) or for Cdc2 kinase activity (measured as histone H1 kinase activity, H1K, in the bottom panel). (C) xTRF1 dissociates from chromatin upon transition from M phase to interphase in egg extracts. Sperm chromatin was incubated with CSF extracts supplemented with 35S-xTRF1 for 10 min and the sample was divided into three parts. One part was further incubated for 60 min, chromatin was isolated at the indicated time points and bound proteins were analyzed as in A (lanes 1–4). The second part was supplemented with Ca2+ to make a final concentration of 0.4 mM (time 0), incubated for 60 min, and chromatin that was isolated at the indicated time points was analyzed. The third part was supplemented with ΔN-cyclin B (ΔN-cycB) 10 min prior to the addition of Ca2+ (time 0), and chromatin was analyzed (lanes 9–12). (D) xTRF1 and xTRF2 specifically recognize telomeric repeat sequence in Xenopus egg extracts. Linear plasmid DNAs harboring ∼800-bp telomeric repeats at one end (pT2AG3) or the control plasmid (control DNA) was conjugated to paramagnetic beads at the nontelomeric end, and incubated with CSF (M) or interphase (I) egg extracts supplemented with 35S-xTRF1 or -xTRF2 for 1 h. Chromatin-bound xTRF1 and xTRF2 were analyzed by SDS–PAGE, followed by phosphorimaging (panels a and c or panels b and d, lanes 4 and 5, respectively). Alternatively, DNA beads were incubated in CSF extracts with 35S-xTRF1 for 20 min, and then Ca2+ and CHX (at final concentrations of 0.6 mM and 100 μg/ml, respectively) were added to induce a transition from M phase to interphase. The chromatin was further incubated for 40 min, and bound proteins were analyzed as described above (lane 6). 35S-xTRF1 and -xTRF2 present in the total extracts are indicated in lanes 1–3 of each panel. (E) xTRF2 is a phosphoprotein. Paramagnetic beads conjugated with pT2AG3 were incubated with CSF (M, lanes 2, 4, 6 and 8) or interphase (I, lanes 3, 5 and 7) egg extracts. After 1 h incubation, the beads were magnetically isolated from the extracts and incubated at 30°C in the presence (lanes 4–7) or absence (lanes 2 and 3) of lambda phosphatase for 2 h. In lanes 6 and 7, phosphatase inhibitors (50 mM NaF, 10 mM orthosodium vanadate and 1 μM OA) were simultaneously added to the reaction. The DNA beads were isolated magnetically again and analyzed by immunoblotting with antibodies against xTRF2. In vitro translated xTRF2 was also analyzed (IVT, lane 1). Download figure Download PowerPoint We next examined the possibility that xTRF1 is recruited to and dissociates from chromatin upon the transitions from interphase to M phase, and from M phase to interphase, respectively. The N-terminally truncated form of cyclin B (ΔN-cyclin B) lacks the destruction box and is constitutively active. The addition of ΔN-cyclin B to interphase extracts leads to the activation of Cdc2 kinase and a transition to mitotic extracts (Glotzer et al, 1991). When recombinant ΔN-cyclin B was added to interphase extracts that had been incubated for 1 h with sperm chromatin and 35S-xTRF1, Cdc2 kinase activity (measured as histone H1 kinase activity) was increased (Figure 1B, bottom panel, lanes 10–13), as expected. Interestingly, xTRF1 became associated with the mitotic chromatin at 20 min and thereafter, indicating that the association of xTRF1 with chromatin is coupled with the transition from interphase to M phase. Conversely, when Ca2+ was added to CSF extracts in which xTRF1 had been incubated with chromatin for 20 min, the bound xTRF1 became dissociated from chromatin 40 min after the Ca2+ addition, following the inactivation of Cdc2 kinase 20 min after the Ca2+ addition (Figure 1C, lanes 5–8; note the 20-min lag between the start of Cdc2 kinase inactivation and that of xTRF1 dissociation). However, the dissociation did not occur when Cdc2 kinase activity was maintained by adding ΔN-cyclin B simultaneously with Ca2+ (lanes 9–12). These results indicate that the dissociation of xTRF1 from chromatin is coupled with the transition from M phase to interphase, and intimately related to the reduced Cdc2 kinase activity. Taken together, we concluded that xTRF1 actively associates with mitotic chromatin and dissociates from interphase chromatin with an apparently strong relationship with cyclin B–Cdc2 kinase activity. xTRF1 and xTRF2 specifically bind to telomeric repeat sequence in Xenopus egg extracts The experiments described above demonstrated that xTRF1 binds to chromatin in a cell-cycle-dependent manner, whereas xTRF2 constitutively associates with chromatin throughout the cell cycle. However, they did not reveal whether the association of xTRF1 and xTRF2 with chromatin is through specific binding to telomeric repeats or nonspecific binding to general chromatin. To this end, linear plasmid DNAs with or without an ∼800-bp telomeric repeat array were conjugated to paramagnetic beads, and incubated with Xenopus egg extracts and 35S-xTRF1 or 35S-xTRF2 for 1 h to allow for chromatin formation on the beads (Sandaltzopoulos and Becker, 1999). Reconstituted chromatin was isolated in a magnetic field and bound proteins were analyzed (Figure 1D). We found that 35S-xTRF1 was specifically recovered from the telomeric DNA beads incubated in CSF extracts, but not from those incubated in interphase extracts, or control DNA beads incubated in CSF or interphase extracts (panels a and c, lanes 4 and 5). Furthermore, when Ca2+ was added to telomeric DNA beads that had been incubated in CSF extracts for 20 min, the bound xTRF1 dissociated from the beads 40 min after the Ca2+ addition (lane 6). 35S-xTRF2 was also detected only in the telomeric DNA beads fraction and its binding activity was not affected by the cell-cycle stage (panels b and d), as observed in the chromatin-binding assay. We found that both 35S-xTRF1 and -xTRF2 proteins were very stable in Xenopus egg extracts even in the absence of sperm chromatin or DNA beads (data not shown), making it unlikely that the different half-lives of xTRF1 and xTRF2 were responsible for the distinct recoveries of the proteins. Moreover, when endogenous xTRF2 was analyzed using anti-xTRF2 antibodies, it behaved in a manner similar to 35S-xTRF2 (Figure 1E, lanes 2 and 3). These results clearly indicate that xTRF1 and xTRF2 specifically recognize and bind to telomeric repeat DNA in Xenopus egg extracts. We noted that endogenous xTRF2 bound to telomeric DNA beads showed different mobility on SDS–PAGE (Figure 1E, lanes 2 and 3), as found in the chromatin-bound fractions (Figure 1A). When these fractions were treated with lambda phosphatase, not only mitotic xTRF2 but also interphase xTRF2 showed an increase in mobility in a manner sensitive to phosphatase inhibitors (lanes 4–7). This faster migrating form of xTRF2 showed the same mobility as the recombinant xTRF2 prepared in RRL (lane 1), and was also detected when CSF extracts were treated with 6-dimethylaminopurine (6-DMAP), an inhibitor of various protein kinases, including cyclin-dependent kinases (Blow, 1993) (Figure 1E, lane 8). Similar results were obtained for endogenous xTRF2 that was immunoprecipitated from CSF and interphase extracts without sperm chromatin (i.e., total xTRF2 present in the extracts, data not shown). We concluded that total xTRF2 in CSF and interphase extracts are hyperphosphorylated and phosphorylated, respectively. In addition, the fact that xTRF2 remains bound to telomeric DNA beads after lambda phosphatase or 6-DMAP treatment suggests that xTRF2 is recruited to telomere independent of phosphorylation. Telomeric DNA binding of chimeric TRF proteins As xTRF1 and xTRF2 lack the amino-terminal acidic or basic region that is present in mammalian TRF1 and 2 proteins, respectively, they have similar domain architecture, consisting of the amino-terminal TRFH domain, the carboxyl-terminal Myb domain and the linker region connecting those two domains. The Myb domain, the TRFH domain and the linker region show high, moderate and least sequence similarities, respectively, between xTRF1 and xTRF2 (Supplementary Figure 1B). The Myb and TRFH domains are involved in DNA binding and protein–protein interactions, including dimer formation, respectively. As xTRF1 and xTRF2 bind to telomere chromatin during the cell cycle in distinct manners, we wished to pinpoint the region responsible for the difference in behavior. A series of chimeric proteins that are composed of the TRFH and Myb domains and the linker region derived from either xTRF1 or xTRF2 in all combinations were constructed (Figure 2A; in the construct names, a set of three numbers following 'TRF' indicates the origin of the three domains; e.g., xTRF211 is a chimera of xTRF2-TRFH, xTRF1-linker and xTRF1-Myb). These proteins were produced as 35S-labeled recombinant proteins in RRL, and examined for their ability to bind to mitotic and interphase telomeric DNA beads (Figure 2B). All the chimeric proteins bound to mitotic telomeric DNA beads as efficiently as the wild-type protein, suggesting that they are functional. The finding that all the chimeric constructs bound to mitotic telomeric DNA beads suggests that binding to mitotic telomeres is a default property of xTRF1 and xTRF2, and that the dissociation of xTRF1 from interphase telomeres is a characteristic specifically conferred to xTRF1. The differential binding to mitotic and interphase telomeric DNA beads revealed in Figure 2B was scored and is summarized in Figure 2A. When the TRFH domain or the linker region of xTRF1 was swapped with the corresponding region of xTRF2 (xTRF211 and xTRF121), the interphase-specific dissociation was abolished, indicating that these two regions are required for the dissociation. In contrast, the xTRF1 Myb domain was dispensable for the dissociation (xTRF112). Substitution of the xTRF1 linker region in xTRF2 (xTRF212) conferred only a weak dissociation of the protein from interphase telomeric DNA. Similarly, xTRF2 with xTRF1-derived TRFH (xTRF122) showed robust binding to interphase telomeric DNA. It is difficult to interpret the results of the TRFH domain swapping because such mutant proteins may dimerize with endogenous xTRF protein having the cognate TRFH domain. Together, the combination of the TRFH domain and the linker region of xTRF1 is required and sufficient for the interphase-specific dissociation of xTRF1 from telomeres. Figure 2.Telomeric DNA binding of chimeric xTRF proteins. (A) Schematic diagram depicting the chimeric xTRF proteins. Domains of xTRF1 and xTRF2 are represented by white and shaded boxes, respectively. The location of each domain is shown on the xTRF1 and xTRF2 structures. Telomere binding measured in mitotic or interphase extracts for each protein was scored from the results in (B) and is shown to the right of the diagrams (binding). The extent of mobility shift of chimeric proteins in mitotic extracts is also summarized (shift in M). (B) Telomeric DNA binding of chimeric xTRF proteins in Xenopus egg extracts. All the chimeric proteins were produced as 35S-labeled proteins in RRLs, and analyzed for telomeric DNA binding as described in Figure 1D. Download figure Download PowerPoint Cdc2 kinase activity is not essential for association of xTRF1 with chromatin One potential explanation for the mechanism of xTRF1 dissociation from chromatin in interphase extracts is that xTRF1 is passively excluded from chromatin during the initiation or progression of DNA replication. However, we found no effect of various DNA replication inhibitors, such as Geminin, p21 and aphidicolin, which inhibit the pre-replicative complex (pre-RC) assembly, DNA polymerase loading and fork progression, respectively, on the chromatin binding of xTRF1 (Supplementary Figure 3). Therefore, xTRF1 is not passively excluded from chromatin by events related to DNA replication. We found that the amount of xTRF1 bound to mitotic chromatin was decreased when CSF extracts were treated with 6-DMAP (Supplementary Figure 3B), suggesting that the chromatin targeting of xTRF1 is regulated by a kinase/phosphatase balance in the extracts. One candidate molecule for regulating the M-phase-specific association of xTRF1 with chromatin is Cdc2 kinase. To investigate this possibility, we prepared cycloheximide (CHX)-treated interphase extracts. Sperm chromatin was first incubated with 35S-xTRF1 in CSF extracts and then Ca2+ and CHX were simultaneously added. In these extracts, endogenous B- and A-type cyclins are degraded during exit from M phase, whereas de novo cyclin synthesis is inhibited by CHX, resulting in complete inactivation of Cdc2 kinase. H1 kinase assay confirmed that no Cdc2 kinase activity was detectable in the CHX-treated interphase extracts (Figure 3A, lane 2). xTRF1 associated with mitotic chromatin (lane 1), but not with interphase chromatin in the presence of CHX (lane 2), as expected. It is known that the addition of okadaic acid (OA), an inhibitor of type1 and type 2A phosphatases, to interphase extracts in the presence of CHX activates M-phase-specific protein kinases other than Cdc2 kinase, thereby inducing a state mimicking M phase, as represented by Cdc25 activation and histone H3 phosphorylation (Sumara et al, 2000). Owing to the inability to synthesize cyclins due to the presence of CHX, Cdc2 kinase is inactive in these extracts (lane 3). Interestingly, we found that xTRF1 associated with chromatin in the OA-induced pseudo-M phase extracts (lane 3). These results indicate that although Cdc2 activation appears to be intimately related to xTRF1 association with chromatin in M phase (Figure 1), it is not essential for the association, as revealed in the OA-induced pseudo-M phase extracts (Figure 3A). Figure 3.Plx1 phosphorylates xTRF1 in vitro. (A) XSP nuclei (4000 nuclei/μl) were incubated for 20 min at 22°C in CSF extracts with 35S-xTRF1, and then Ca2+ and CHX (100 μg/ml) were simultaneously added to release meiotic metaphase II arrest in the presence (lane 3) or absence (lane 2) of 1 μM OA. As control, the extracts were incubated without the addition of Ca2+ or CHX (lane 1). After 1 h incubation, total extracts (top panel) and chromatin-bound proteins (second panel) were analyzed as described in Figure 1B. (B) Immunoblot analysis of CSF extracts with antibodies against Plx1 or Xenopus Aurora B. The antibodies were raised against the carboxyl-terminal peptide sequences and affinity-purified. (C) Immunoprecipitation of Plx1 or Xenopus Aurora B. CSF extracts (lane 1) were incubated with normal rabbit IgG (lane 2), anti-Plx1 antibodies (lane 3) or anti-Xenopu
Referência(s)