Geminin Becomes Activated As an Inhibitor of Cdt1/RLF-B Following Nuclear Import

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Current Biology, Vol. 12, 678–683, April 16, 2002, 2002 Elsevier Science Ltd. All rights reserved. PII S0960-9822(02)00778-9 Geminin Becomes Activated as an Inhibitor of Cdt1/RLF-B Following Nuclear Import

Ben Hodgson,2 Anatoliy Li, Shusuke Tada,3 is released from the nuclei, so the PIEs contain a mixture and J. Julian Blow1 of both cytoplasmic and nucleoplasmic proteins (data Cancer Research UK Chromosome Replication not shown). When DNA was present during the preincu- Research Group bation, the resultant PIE contained a potent inhibitor of Wellcome Trust Biocentre replication licensing (Figure 1A) and blocked the loading University of Dundee of Mcm3 onto chromatin (Figure 1B). In contrast, when Dow Street no DNA was present in the preincubation, no licensing Dundee DD1 5EH inhibitor was generated. The PIE had no inhibitory effect United Kingdom on the replication of licensed DNA templates, suggesting that the inhibitory activity is specific for licensing (data not shown). Summary Since the inhibitory activity first appeared around the time that nuclear assembly was completed (data not During late mitosis and early interphase, origins of shown), we investigated whether nuclear assembly is replication become “licensed” for DNA replication by involved in its generation. Ran T24N, a mutant Ran pro- loading Mcm2-7 complexes [1–5]. Mcm2-7 complexes tein stabilized predominantly in the GDP-bound form, are removed from origins as replication forks initiate inhibits nuclear assembly in Xenopus extract [12, 13]. replication, thus preventing rereplication of DNA in a Figure 1A shows that, when extract was preincubated single cell cycle. Premature origin licensing is pre- for 90 min with sperm nuclei and Ran T24N protein, no vented in metaphase by the action of geminin, which licensing inhibitor was generated. In contrast, wild-type binds and inhibits Cdt1/RLF-B, a protein that is re- Ran (which does not block nuclear assembly) did not quired for the loading of Mcm2-7 [6–9]. Recombinant block the appearance of the inhibitor. Wheat germ ag- geminin that is added to Xenopus egg extracts is effi- glutinin, a lectin that binds to nuclear pores, inhibits ciently degraded upon exit from metaphase [10]. Here, nuclear protein transport and nuclear envelope assem- we show that recombinant and endogenous forms of bly [14]. No licensing inhibitor was generated when DNA Xenopus geminin behave differently from one another, was preincubated in extract supplemented with wheat such that a significant proportion of endogenous gem- germ agglutinin (Figure 1A). The concentration of wheat inin escapes proteolysis upon exit from metaphase. germ agglutinin used in this experiment was sufficient During late mitosis and early G1, the surviving popula- to inhibit DNA replication and gross nuclear swelling but tion of endogenous geminin does not associate with did not block complete nuclear envelope assembly (data Cdt1/RLF-B and does not inhibit licensing. Following not shown), suggesting that inhibition of nuclear trans- nuclear assembly, geminin is imported into nuclei and port was responsible for the effect. Similarly, when becomes reactivated to bind Cdt1/RLF-B. This reacti- sperm nuclei were incubated in extract lacking nuclear vated geminin provides the major nucleoplasmic in- envelope precursors (membrane-free extract), no licens- hibitor of origin relicensing during late interphase. ing inhibitor was generated (Figure 1A). Taken together, Since the initiation of replication at licensed origins these results suggest that nuclear assembly and nuclear depends on nuclear assembly [11], our results suggest transport are required for generation of the inhibitory an elegant and novel mechanism for preventing rerep- activity in preincubated extracts. lication of DNA in a single cell cycle. We next investigated whether the inhibitory activity is localized in the nucleus. Nucleoplasmic and cytoplasmic Results fractions of Xenopus egg extracts can be prepared by floating intact nuclei through the cytoplasm by gentle DNA-Dependent Generation of an Inhibitor centrifugation [15]. Hard centrifugation of the nuclear of Replication Licensing fraction then separates it into a soluble nucleoplasmic As part of our studies into how relicensing of replicated extract and an insoluble pellet. When preincubated ex- DNA is prevented in G2, we investigated whether the tract was fractionated in this way, most of the licensing presence of DNA in Xenopus extract affected further inhibitor present in preincubated extracts was seen to replication licensing. Extract was supplemented with be concentrated in the nucleoplasmic extract (Figure cycloheximide (to arrest extracts in G2 phase) plus or 1C). It was of interest that Cdk2 and geminin (both poten- minus sperm nuclei and was incubated for 90 min. Ex- tial licensing inhibitors) were also concentrated within tract was then diluted and the chromatin was removed the nucleoplasmic extract (Figure 1D). by centrifugation to generate a “preincubated extract” (PIE). In this protocol, almost all of the soluble protein The Licensing Inhibitor Is Geminin 1Correspondence: [email protected] Studies in yeast have shown that high cyclin-dependent 2 Present address: Paterson Institute for Cancer Research, Christie kinase (CDK) activity is important for preventing reli- Hospital NHS Trust, Wilmslow Road, Manchester M20 4BX, United Kingdom. censing of DNA and reassembly of prereplicative com- 3 Present address: Graduate School of Pharmaceutical Sciences, plexes (pre-RCs) late in the cell cycle [2, 3]. To investi- Tohoku University, Aoba-ku, Sendai, Miyagi 980-8578, Japan. gate whether our licensing inhibitor depended on high Brief Communication 679

Figure 1. DNA-Dependent Generation of an Inhibitor of Replication Licensing (A) Whole extract or membrane-free extract was preincubated with or without sperm nuclei (10 ng DNA/␮l) Ϯ 2.5 ␮M wild-type or T24N mutant Ran or 200 ␮g/ml wheat germ agglutinin. The ability of the PIEs to inhibit licensing in untreated interphase extracts was measured. The dashed line represents background licensing of sperm nuclei incubated in buffer alone. (B) Sperm nuclei were incubated for 30 min in interphase extract supplemented with either buffer or PIE. Chromatin was isolated and immunoblotted for Orc1, Cdc6, Mcm3, Cdt1, and geminin. (C) PIE, nucleoplasmic extract, and cytoplasmic extract were subject to serial 3-fold dilutions, then assayed for their ability to inhibit licensing of interphase extract. The dashed line represents background licensing of sperm nuclei incubated in buffer alone. (D) Cytoplasmic extract, nucleoplasmic extract, and the insoluble nuclear pellet were immunoblotted for Orc1, Cdc6, Cdt1, Mcm7, geminin, and CDKs (PSTAIR).

CDK activity, we used the specific CDK inhibitor roscovi- tory activity generated in G2 as a result of nuclear as- tine [16]. Although 2.5 mM roscovitine is sufficient to sembly and nuclear transport. inhibit the initiation of DNA replication [16], it did not We were surprised to find geminin in our extracts, block the appearance of the licensing inhibitor (Figure since geminin was originally identified in a screen for 2A, PIE ϩ rosc.). Similar results were obtained with the proteins specifically degraded at the metaphase-to- CDK inhibitor p21Cip1 (data not shown). Interphase ex- anaphase transition [10] and our extracts were supple- tracts were depleted of ORC, Cdc6, Cdt1, Mcm3, Mcm7, mented with cycloheximide to prevent any new protein or Cdc7, or were mock depleted with nonimmune serum. synthesis. We noted, however, that the assays per- When DNA was preincubated in each of these extracts, formed by McGarry and Kirschner [10] involved the deg- a potent licensing inhibitor was still generated (Figure 2A), radation of recombinant geminin, while the degradation suggesting that generation of the licensing inhibitor is not of endogenous geminin was not reported. To confirm dependent on any known aspect of pre-RC assembly. that our extracts could efficiently degrade recombinant The major inhibitor of origin licensing present in meta- geminin upon exit from metaphase, two forms of 6xHis- phase Xenopus egg extract is geminin [9]. Figure 2A tagged geminin protein were produced in Escherichia shows that interphase extracts immunodepleted of coli: one wild-type (Figure 2B, lane 1), and one con- geminin failed to generate the licensing inhibitor after taining a mutation in its destruction box to make it resis- preincubation with DNA. In addition, when inhibitory tant to degradation (gemininDEL, Figure 2B, lane 4) [10]. PIEs prepared from untreated extract were subse- When incubated in extracts exiting from metaphase, the quently depleted of geminin, the inhibitory activity was wild-type recombinant protein was almost completely lost (Figure 2A, PIE, gemininϪ). The addition of recombi- degraded, while the DEL mutant remained stable (Figure nant Cdt1 (the target of geminin inhibition) to PIE reacti- 2B, lanes 2 and 5). In contrast, both proteins remained vated its ability to assemble Mcm7 onto sperm chroma- stable when incubated for an equivalent period of time tin (Figure 2A, inset). When similar experiments were in interphase extract (Figure 2B, lanes 3 and 6). These performed on nucleoplasmic fractions, immunodeple- results are consistent with the efficient cell cycle-spe- tion of geminin removed the licensing inhibitor, while cific degradation of recombinant geminin reported by inhibition of CDKs did not (Figure S1). These results McGarry and Kirschner [10]. therefore show that geminin provides the major inhibi- However, in these experiments, significant quantities Current Biology 680

Figure 2. The Licensing Inhibitor Is Geminin (A) The ability of various preincubated extracts to inhibit replication licensing was assayed. Sperm nuclei were preincubated in interphase extract Ϯ 2.5 mM roscovitine (PIE, PIE ϩ rosc.). Alternatively, sperm nuclei were preincubated in interphase extract previously depleted with antibodies against Orc1, Cdc6, Mcm3, Mcm7, Cdc7, or geminin, or mock depleted with nonimmune antibodies (NIϪ), or extract previously depleted of Cdt1 using geminin-coupled beads. Alternatively, PIE was prepared in normal interphase extract and was then depleted with either anti-geminin antibodies (PIE, gemininϪ) or nonimmune antibodies (PIE, NIϪ). Inset: immunoblot of chromatin-bound Mcm7 following 30 min of incubation of sperm nuclei in PIE (lane 1), PIE ϩ recombinant Cdt1 (lane 2), or PIE prepared without DNA (lane 3). (B) Wild-type and gemininDEL (which lacks a functional destruction box) were incubated at 1.3 ng/␮l for 45 min in metaphase extract

released into interphase with CaCl2 (lanes 2 and 5) or interphase extract (lanes 3 and 6). Input levels of recombinant geminin are shown in lanes 1 and 4. Samples were analyzed by SDS-PAGE and immunoblotted with anti-geminin antibodies (top panel) or anti- 6xHis antibodies (bottom panel). (C) Metaphase extract (M) Ϯ sperm nuclei (10 ng DNA/␮l) was released into interphase by

the addition of CaCl2, and, at the indicated times (in min), aliquots were taken and immunoblotted for endogenous geminin and Cdt1.

,kDa. Geminin was observed in two peaks 500ف of endogenous geminin remained in the extract (Figure peak of ,in a large peak cofractionating with Cdt1/RLF-B %61ف -2B, upper panel, lanes 2 and 3). The quantity of endoge nous geminin and Cdt1/RLF-B present in extracts was and the rest in a smaller peak without Cdt1/RLF-B (Fig- therefore measured at different times after exit from ure 3A). The smaller peak migrated in the same position metaphase arrest in vitro (Figure 2C). Although some as free recombinant geminin (data not shown). This sug- geminin degradation was clearly observed upon exit gests that, in metaphase, Cdt1/RLF-B is fully bound from metaphase, 30%–60% of endogenous geminin re- and inhibited by geminin and that there is also excess sisted degradation. We also noticed a slight decline in geminin activity present [9]. Upon exit from metaphase the amount of Cdt1/RLF-B, consistent with the loss of into interphase, some, but not all, geminin was de- RLF-B activity seen when preincubations were per- graded, but Ͼ95% of the remaining geminin was found formed without added DNA [17]. The presence of sperm in the smaller complex (Figure 3B). Further, the endoge- nuclei in the extract had no significant effect on the nous Cdt1/RLF-B migrated as a slightly smaller com- degree of degradation observed (Figure 2C). We con- plex, consistent with it no longer binding geminin. When clude that a significant quantity of endogenous geminin extract was preincubated with DNA, so that geminin escapes degradation, and that this geminin is incapable once again inhibited Cdt1/RLF-B, the fractionation pat- of inhibiting licensing until nuclear assembly occurs. tern again resembled that seen in metaphase, with all geminin cofractionating in %45ف the Cdt1/RLF-B and Geminin’s Inhibitory Activity Correlates with Its the larger complex (Figure 3C). Immunodepletion experi- Ability to Interact with Cdt1/RLF-B ments showed that, under these conditions, a complex If some endogenous geminin escapes degradation in between Cdt1/RLF-B and geminin had reformed (Figure anaphase, why does it not inhibit licensing? A change S2). In contrast, when extract was preincubated in the in its ability to interact with Cdt1/RLF-B could provide absence of DNA, most geminin and Cdt1/RLF-B did of the %24ف) a potential explanation. We therefore examined the be- not cofractionate, though some complex havior of geminin and Cdt1/RLF-B by gel filtration. In geminin) was formed (Figure 3D). Coprecipitation also metaphase-arrested extracts, Cdt1 migrated in a single showed a significant quantity of free Cdt1 in the absence Brief Communication 681

Figure 3. Geminin’s Affinity for Cdt1 Changes during the Cell Cycle (A–D) Gel filtration of different extracts immunoblotted for geminin and Cdt1: (A) metaphase extract, (B) interphase extract (10 min after metaphase release), (C) preincubated extract plus sperm (10 ng DNA/␮l, 90 min), and (D) preincubated extract minus sperm (no DNA, 90 min). The migration of molecular weight markers (in kDa) is shown above. (E) Preincubated extracts were prepared Ϯ 10 ng DNA/␮l at the indicate times; 6xHis- Cdt1 was added and incubated for 30 min, then pulled out on Ni-NTA agarose. Bead- bound Cdt1 and geminin were analyzed by SDS-PAGE and immunoblotting.

of DNA (Figure S2). We therefore see a good correspon- Discussion dence between the presence of inhibitory activity and the cofractionation of Cdt1/RLF-B and geminin. Replication licensing requires ORC to bind origin DNA, Since recombinant geminin can bind Cdt1/RLF-B and which in turn permits the recruitment of Cdc6 and Cdt1/ inhibit licensing when added to early interphase Xeno- RLF-B, and eventually the loading of Mcm2-7 [1–5]. pus extracts [9, 10], we reasoned that endogenous gemi- While each of the different components can potentially nin might become altered during late mitosis and early be regulated [5], previous studies have indicated that interphase so that it is unable to bind Cdt1/RLF-B. To origin licensing is controlled within the Xenopus embry- address this, extract was preincubated plus or minus onic cell cycle, predominantly by the regulation of Cdt1/ DNA for either 5 or 90 min. Recombinant 6xHis-Cdt1 RLF-B activity [9, 17, 18]. Cdt1/RLF-B is specifically was added to these PIEs and was then collected on Ni- bound and inhibited by geminin, and this provides the agarose beads. Bead-bound material was immunoblot- major mechanism by which origin licensing is regulated ted to determine whether endogenous geminin had during metaphase [9]. Geminin was originally identified bound to the recombinant Cdt1. Figure 3E shows that, in a screen for recombinant proteins degraded upon exit when nuclear assembly did not occur (either because from mitosis in Xenopus egg extracts [10]. However, no DNA was present or because insufficient time had this earlier study did not investigate the degradation of elapsed to allow nuclei to assemble), endogenous gemi- endogenous geminin in the Xenopus system. We dem- nin did not bind significantly to the recombinant Cdt1. onstrate here that, although recombinant geminin is effi- However, when nuclear assembly occurred and the in- ciently degraded by Xenopus egg extracts upon exit hibitor was activated, a significant quantity of endoge- from metaphase, a significant proportion of endogenous nous geminin was bound to the recombinant Cdt1 (Fig- geminin escapes degradation. This is different from ure 3E, lane 5). This suggests that, as a consequence what is observed in somatic cells, where geminin ap- of nuclear assembly, endogenous geminin becomes pears to be virtually absent throughout G1 [10, 19]. competent for inhibition of Cdt1/RLF-B. Despite the continued presence of endogenous gemi- Current Biology 682

tively, there may be another protein that is activated upon exit from metaphase that modulates the activity of geminin in a noncovalent way. We also show that endogenous geminin regains the capacity to bind recombinant Cdt1 and inhibit further origin licensing when it has been imported into an interphase nucleus (Figure 4C). It is possible that the factor(s) required to inactivate geminin are excluded from interphase nuclei, thus allowing the alteration that had occurred upon metaphase exit to become reversed in- side the nucleus. The reactivation of geminin did not depend on CDK activity, distinguishing this mechanism from what has been reported in yeast, where high nuclear CDK activity is required to prevent reassembly of pre-RCs and relicensing of replicated DNA [2, 3, 5]. This parallels earlier work showing that geminin is the major inhibitor of origin licensing in metaphase-arrested Xeno- pus egg extracts [9]. Since the initiation of DNA replication only takes place within the nucleus [11], the reactivation of geminin as a licensing inhibitor in the nucleus provides an elegant mechanism for preventing rereplication of DNA.

Experimental Procedures

Preparation of Egg Extracts and Sperm Nuclei Metaphase-arrested and interphase low-speed supernatant Xeno- Figure 4. Model for Geminin Regulation during the Xenopus Embry- pus egg extracts were prepared as described [21]. Extracts were onic Cell Cycle supplemented with 250 ␮g/ml cycloheximide, 25 mM phosphocre- ␮ A cartoon showing a small segment of chromatin containing a repli- atine, and 10 g/ml creatine phosphokinase before use, and incuba- Њ cation origin as it passes through the cell cycle. Hexagons labelled tions were performed at 23 . 6-DMAP-treated extract was prepared “M” represent hexameric complexes of Mcm2-7. by supplementing metaphase-arrested extracts with 3 mM 6-DMAP, ␣ 32 (A) In metaphase, geminin is both abundant and active, so it is able [ - P]dATP, and 0.3 mM CaCl2 [20]. Xenopus sperm nuclei were to inhibit origin licensing by formation of a tight complex with Cdt1/ demembranated with lysolecithin as described [21] and frozen in RLF-B. ORC and Cdc6 do not bind tightly to DNA, probably due to aliquots in liquid nitrogen. Preincubated extracts (PIEs) were pre- high CDK levels. pared by incubating interphase extract with sperm nuclei at 10 ng ␮ (B) Upon metaphase release, some of the geminin is degraded, but DNA/ l for 90 min. Extracts were then diluted by the addition of 3 the remaining geminin is altered in some way that prevents it from vol ice-cold LFB2/50 (40 mM HEPES KOH [pH 8.0], 50 mM KCl, 20 binding and inhibiting Cdt1/RLF-B. The loss of CDK activity allows mM K2HPO4/KH2PO4 [pH 8.0], 2 mM MgCl2, 1 mM EGTA, 2 mM DTT, ␮ ORC and Cdc6 to associate tightly with DNA. With all components 2.5 mM Mg-ATP, 10% sucrose, and 1 g/ml each of leupeptin, of the licensing system active, Mcm2-7 is loaded onto origins. pepstatin, and aprotinin). The diluted extracts were then transferred (C) Following nuclear assembly, geminin is imported into a functional to precooled polycarbonate centrifuge tubes (Beckman #343775) ϫ ف nucleus and becomes reactivated. This allows it to bind and inhibit and centrifuged at 72,000 rpm ( 30,000 g) using a TLA-100 rotor Њ Cdt1/RLF-B, thus preventing the relicensing of replicated origins. (Beckman) for 20 min at 4 C. The supernatant was recovered and snap frozen in 10 ␮l beads in liquid nitrogen. Membrane-free extracts were prepared by 3-fold dilution of ex- nin, Cdt1/RLF-B activity appears abruptly upon exit from tract in LFB2/50, followed by centrifugation at 72,000 rpm .ϫ g) using a TLA-100 rotor (Beckman) for 20 min at 4ЊC 230,000ف) metaphase [17, 20]. We show here that, as Cdt1/RLF-B becomes activated upon exit from metaphase, it changes Nucleoplasmic extract and cytoplasmic extract were prepared as previously described [15]. Immunodepletion of interphase extracts, from being in a complex with geminin to a form where PIE, and nucleoplasmic extract with antibodies raised against XOrc1 it is not. When recombinant geminin is added to early [22], XCdc6 [18], XMcm3 [23], XMcm7 [24], geminin [9], or with interphase Xenopus egg extracts, it forms a complex antibodies from nonimmune rabbit serum was performed as de- with endogenous Cdt1/RLF-B and strongly inhibits li- scribed [21]. Extract was depleted of Cdt1 using geminin-coupled censing [9, 10]. In contrast, we show here that, when beads as described [9]. recombinant Cdt1/RLF-B is added to the extract, it binds endogenous geminin only very poorly. This suggests Licensing Assays that, upon exit from metaphase, geminin is altered in Licensing inhibition assays were performed essentially as described [21]. Briefly, 2-␮l fractions of interest were incubated with 0.4 ␮l some way that prevents it binding to and inhibiting Cdt1/ interphase extract and 0.3 ␮l sperm nuclei (80 ng DNA/␮l) for 30 RLF-B. This change leads to the abrupt activation of min. The degree of licensing was then assessed by the addition of origin licensing upon exit from metaphase (Figures 4A 6 ␮l 6-DMAP-treated extract containing [␣-32P]dATP and incubation and 4B). It is possible that this geminin alteration is also for 90 min; total DNA synthesis was measured by scintillation count- responsible for protecting it from degradation. Although ing of acid-insoluble material. we have detected different forms of geminin on two- To investigate chromatin binding, licensing assays were scaled up 10-fold to a total volume of 24 ␮l. After 30 min of incubation, dimensional gels, we have so far been unable to identify each reaction was diluted in 300 ␮l NIBA (50 mM KCl, 50 mM HEPES

a covalent modification that strictly correlates with the KOH [pH 7.6], 5 mM MgCl2, 2 mM DTT, 0.5 mM spermidine, 0.15 changes in activity (A.L., unpublished data). Alterna- mM spermine, 2.5 mM Mg-ATP, and 1 ␮g/ml each of leupeptin, Brief Communication 683

pepstatin, and aprotinin) supplemented with 0.1% Triton X-100 and 3. Lei, M., and Tye, B.K. (2001). Initiating DNA synthesis: from underlayered with 100 ␮l of the same buffer containing 15% sucrose. recruiting to activating the MCM complex. J. Cell Sci. 114, 1447– The chromatin was pelleted at 3,500 ϫ g in a swinging bucket rotor 1454. at 4ЊC. The diluted extract was removed, and the chromatin pellet 4. Gillespie, P.J., Li, A., and Blow, J.J. (2001). Reconstitution of was resuspended in SDS-PAGE loading buffer. licensed replication origins on Xenopus sperm nuclei using purified proteins. BMC Biochem. 2, 15. Recombinant Proteins, Immunoblotting, 5. Blow, J.J., and Hodgson, B. (2002). Replication licensing - defin- and Pull-Down Assays ing the proliferative state? Trends Cell Biol. 12, 72–76. Recombinant 6xHis-tagged Xenopus geminin and gemininDEL [10] 6. Maiorano, D., Moreau, J., and Mechali, M. (2000). XCDT1 is were produced from expression plasmids (kindly provided by Tom required for the assembly of pre-replicative complexes in Xeno- McGarry) and were purified on Ni-agarose by standard techniques. pus laevis. Nature 404, 622–625. Recombinant 6xHis-tagged Xenopus ⌬Cdt1 was produced as de- 7. Nishitani, H., Lygerou, Z., Nishimoto, T., and Nurse, P. (2000). scribed [4, 6]. To investigate the interaction between geminin and The Cdt1 protein is required to license DNA for replication in Cdt1/RLF-B, recombinant Cdt1 was incubated in mock and preincu- fission yeast. Nature 404, 625–628. bated extracts at room temperature for 30 min. Ni-agarose beads 8. Wohlschlegel, J.A., Dwyer, B.T., Dhar, S.K., Cvetic, C., Walter, (Qiagen), washed extensively in PDB ϩ 66 mM KCl (PDB: 40 mM J.C., and Dutta, A. (2000). Inhibition of eukaryotic replication by geminin binding to Cdt1. Science 290, 2309–2312. HEPES KOH [pH 8.0], 20 mM K HPO /KH PO4 [pH 8.0], 2 mM 2 4 2 9. Tada, S., Li, A., Maiorano, D., Mechali, M., and Blow, J.J. (2001). ␤-mercaptoethanol, 10% (w/v) sucrose, 10 ␮g/ml each of leupeptin, Repression of origin assembly in metaphase depends on inhibi- pepstatin, and aprotinin, 20 mM imidazole, and 0.01% Triton X-100), tion of RLF-B/Cdt1 by geminin. Nat. Cell Biol. 3, 107–113. were then added to extracts and incubated for an additional 30 min 10. McGarry, T.J., and Kirschner, M.W. (1998). Geminin, an inhibitor at 4Њ. Beads were removed by centrifugation through a 25-␮m nylon of DNA replication, is degraded during mitosis. Cell 93, 1043– filter, washed twice with 10 vol PDB ϩ 500 mM KCl, and washed 1053. once with 10 vol PDB ϩ 10 mM KCl. Beads were boiled in SDS 11. Blow, J.J. (2001). Control of chromosomal DNA replication in loading buffer and centrifuged, and the supernatant was immu- the early Xenopus embryo. EMBO J. 20, 3293–3297. noblotted for geminin or Cdt1. 12. Dasso, M., Seki, T., Azuma, Y., Ohba, T., and Nishimoto, T. For immunoblotting, samples were run on 4%–12% gradient No- (1994). A mutant form of the Ran/TC4 protein disrupts nuclear vex precast gels and were analyzed by immunoblotting by ECL function in Xenopus laevis egg extracts by inhibiting the RCC1 detection (Amersham). Antibodies against XOrc1 [22], XCdc6 [18], protein, a regulator of chromosome condensation. EMBO J. 13, XMcm3 [23], XMcm7 [24], XCdt1 [9], geminin [9], 6xHis (Santa Cruz 5732–5744. Biotechnology), and PSTAIR (Zymed) were used. Film was scanned 13. Hughes, M., Zhang, C.M., Avis, J.M., Hutchison, C.J., and and imported into Adobe PhotoShop for printing. Clarke, P.R. (1998). The role of the Ran GTPase in nuclear assembly and DNA replication: characterisation of the effects of Gel Filtration Ran mutants. J. Cell Sci. 111, 3017–3026. A total of 20 ␮l metaphase extract or interphase extract and 50 ␮l 14. Finlay, D.R., Newmeyer, D.D., Price, T.M., and Forbes, D.J. PIE (90 min, Ϯ 10 ng DNA/␮l) were applied to a 2.4-ml Superose 6 (1987). Inhibition of in vitro nuclear transport by a lectin that column (Amersham-Pharmacia) equilibrated in T’LFB1/200 (40 mM binds to nuclear-pores. J. Cell Biol. 104, 189–200.

HEPES KOH [pH 8.0], 200 mM KCl, 20 mM K2HPO4/KH2PO4 [pH 8.0], 15. Walter, J., Sun, L., and Newport, J. (1998). Regulated chromo-

2 mM MgCl2, 1 mM EGTA, 2 mM DTT, 10% sucrose, 1 ␮g/ml each somal DNA replication in the absence of a nucleus. Mol. Cell 1, of leupeptin, pepstatin, and aprotinin, and 0.01% Triton-X 100). The 519–529. column was run at a flow rate of 25 ␮l/min on a SMART chromatogra- 16. Meijer, L., Borgne, A., Mulner, O., Chong, J.P., Blow, J.J., Ina- phy system (Amersham-Pharmacia) at 4ЊC, and 50-␮l fractions were gaki, N., Inagaki, M., Delcros, J.G., and Moulinoux, J.P. (1997). collected. Biochemical and cellular effects of roscovitine, a potent and selective inhibitor of the cyclin-dependent kinases cdc2, cdk2 Supplementary Material and cdk5. Eur. J. Biochem. 243, 527–536. Supplementary Material including two supplementary figures is 17. Mahbubani, H.M., Chong, J.P., Chevalier, S., Tho¨ mmes, P., and available at http://images.cellpress.com/supmat/supmatin.htm. Blow, J.J. (1997). Cell cycle regulation of the replication licens- Figure S1 shows that immunodepletion of geminin, but not inhibition ing system: involvement of a Cdk-dependent inhibitor. J. Cell of CDKs with 6-DMAP, eliminates the ability of nucleoplasmic ex- Biol. 136, 125–135. tracts to inhibit replication licensing. Figure S2 shows coimmuno- 18. Tada, S., Chong, J.P.J., Mahbubani, H.M., and Blow, J.J. (1999). precipitation of geminin and Cdt1 in late interphase only after nuclear The RLF-B component of the replication licensing system is assembly, and not in early interphase. distinct from Cdc6 and functions after Cdc6 binds to chromatin. Curr. Biol. 9, 211–214. 19. Quinn, L.M., Herr, A., McGarry, T.J., and Richardson, H. (2001). Acknowledgments The Drosophila Geminin homolog: roles for Geminin in limiting DNA replication, in anaphase and in neurogenesis. Genes Dev. We would like to thank Paul Clarke for recombinant Ran protein; 15, 2741–2754. Pedro Jares and Peter Gillespie for helpful discussions; and Anne 20. Blow, J.J. (1993). Preventing re-replication of DNA in a single Donaldson, Christine Gosden, Tom Owen-Hughes, and Jason cell cycle: evidence for a replication licensing factor. J. Cell Swedlow for comments on the manuscript. B.H. is the recipient of Biol. 122, 993–1002. a UBI (Upstate Biotechnology) studentship. This work was sup- 21. Chong, J.P., Tho¨ mmes, P., Rowles, A., Mahbubani, H.M., and ported by the Cancer Research UK (CRC) grant SP2385/0101. Blow, J.J. (1997). Characterization of the Xenopus replication licensing system. Methods Enzymol. 283, 549–564. Received: January 23, 2002 22. Rowles, A., Chong, J.P., Brown, L., Howell, M., Evan, G.I., and Revised: February 25, 2002 Blow, J.J. (1996). Interaction between the origin recognition Accepted: February 26, 2002 complex and the replication licensing system in Xenopus. Cell Published: April 16, 2002 87, 287–296. 23. Tho¨ mmes, P., Kubota, Y., Takisawa, H., and Blow, J.J. (1997). References The RLF-M component of the replication licensing system forms complexes containing all six MCM/P1 polypeptides. EMBO J. 1. Blow, J.J., and Laskey, R.A. (1988). A role for the nuclear enve- 16, 3312–3319. lope in controlling DNA replication within the cell cycle. Nature 24. Prokhorova, T.A., and Blow, J.J. (2000). Sequential MCM/P1 332, 546–548. subcomplex assembly is required to form a heterohexamer with 2. Diffley, J.F.X. (2001). Building the perfect switch. Curr. Biol. 11, replication licensing activity. J. Biol. Chem. 275, 2491–2498. R367–370.