Phosphorylation and Activation of 13S Condensin by Cdc2 in Vitro

See allHide authors and affiliations

Science  16 Oct 1998:
Vol. 282, Issue 5388, pp. 487-490
DOI: 10.1126/science.282.5388.487


13S condensin is a multisubunit protein complex essential for mitotic chromosome condensation in Xenopus egg extracts. Purified 13S condensin introduces positive supercoils into DNA in the presence of topoisomerase I and adenosine triphosphate in vitro. The supercoiling activity of 13Scondensin was regulated by mitosis-specific phosphorylation. Immunodepletion, in vitro phosphorylation, and peptide-mapping experiments indicated that Cdc2 is likely to be the kinase that phosphorylates and activates 13S condensin. Multiple Cdc2 phosphorylation sites are clustered in the carboxyl-terminal domain of the XCAP-D2 (Xenopus chromosome-associated polypeptide D2) subunit. These results suggest that phosphorylation of 13Scondensin by Cdc2 may trigger mitotic chromosome condensation in vitro.

Chromosome condensation is a fundamental cellular process that ensures the faithful segregation of genetic information during mitosis and meiosis. Activation of the protein kinase Cdc2 triggers a series of downstream mitotic events including chromosome condensation, but the underlying molecular mechanisms are poorly understood (1, 2). 13S condensin, a five-subunit protein complex purified from Xenopus egg extracts, is an essential regulator of mitotic chromosome condensation (3,4). The two core subunits of 13S condensin, XCAP-C and XCAP-E, belong to the SMC (structural maintenance of chromosomes) family of chromosomal adenosine triphosphatases (ATPases) (2, 5). The remaining three subunits, XCAP-D2, XCAP-G, and XCAP-H, may have regulatory roles in condensin function (4). Genetic studies in yeasts, Drosophila, andCaenorhabditis elegans show that (at least some of) the condensin subunits are essential for chromosome condensation and segregation in vivo (6). When purified from mitotic extracts, 13S condensin has a DNA-stimulated ATPase activity and can introduce positive supercoils into relaxed circular DNA in the presence of ATP and topoisomerase I. This activity may contribute to chromosome condensation during mitosis (7).

To test whether the positive supercoiling activity of 13Scondensin is regulated during the cell cycle, we purified 13S condensin from mitotic or interphase extracts ofXenopus eggs by immunoaffinity column chromatography (8). The subunit compositions of the two forms were indistinguishable, although three of the five subunits (XCAP-D2, XCAP-G, and XCAP-H) were phosphorylated in a mitosis-specific manner (Fig. 1A) (4). Because of this modification, the electrophoretic mobility of the mitotic form of XCAP-H was decreased (Fig. 1A). When this mitotic form of 13S condensin was incubated with a relaxed circular DNA in the presence of ATP and topoisomerase I (purified from calf thymus or Escherichia coli), the DNA was converted into a ladder of supercoiled forms in a dose-dependent manner (Fig. 1B) (9). In contrast, the supercoiling activity was barely detectable in the interphase form of 13S condensin, although it exhibited a DNA binding activity comparable to that of the mitotic form of 13S condensin (Fig. 1B, bottom). Two-dimensional gel electrophoresis confirmed that the mitotic condensin induced positive supercoiling (Fig. 1C). Changes in the average linking number of the substrate DNA were measured to be +3.3 with the mitotic form and +0.1 with the interphase form under this condition (9). The supercoiling activity was highly reproducible between different preparations. Treatment of the mitotic condensin with λ protein phosphatase (λ-PPase) resulted in a decrease in supercoiling activity accompanied by dephosphorylation of the three subunits (10), which suggested that the activity is regulated by mitosis-specific phosphorylation.

Figure 1

Mitosis-specific and phosphorylation-dependent supercoiling activity of 13Scondensin. (A) Characterization of 13Scondensin purified from a mitotic (M) or an interphase (I) extract (8). CBB, Coomassie stain; [32P], autoradiography of condensin subunits purified from32P-labeled extracts; C, XCAP-C; D2, XCAP-D2; E, XCAP-E; G, XCAP-G; H, XCAP-H; H-blot, immunoblot with anti–XCAP-H. (B) Supercoiling and DNA binding activities of the mitotic (lanes 2 to 4) or interphase (lanes 6 to 8) form of 13S condensin (cond.). Calf thymus (top) or E. coli (middle) topoisomerase I was supplemented into the supercoiling reactions but was omitted in the DNA binding assay (bottom) (7). DNA was purified (top and middle) or unpurified (bottom), electrophoresed on a 0.7% agarose gel, and visualized by Southern blotting (9). The molar ratios of 13S condensin to DNA in the reaction mixtures were ∼9:1 (lanes 2 and 6), ∼18:1 (lanes 3 and 7), or ∼36:1 (lanes 4 and 8). Lanes 1 and 5, no protein; s, positively supercoiled DNA; arrow, free DNA; asterisk, DNA bound to 13S condensin. (C) Two-dimensional gel electrophoresis. Substrate DNA (a mixture of nicked circular and relaxed circular DNA) was subjected to the supercoiling assay withE. coli topoisomerase I, fractionated on a two-dimensional agarose gel, and visualized by Southern blotting (9). Left, no condensin; center, mitotic 13S condensin; right, interphase 13S condensin. The molar ratio of protein to DNA was ∼36:1. Abbreviations: nc, nicked circular DNA; rc, relaxed circular DNA; s, positively supercoiled DNA; ns, expected position where negatively supercoiled DNA migrates.

In an attempt to identify the kinase or kinases that activate the supercoiling activity of 13S condensin, we immunodepleted Cdc2 from a mitotic extract (11). The efficiency of immunodepletion was estimated to be >95% by both immunoblotting and measurement of histone H1 kinase activity (Fig. 2A). Depletion of Cdc2 resulted in reduced phosphorylation of the condensin subunits and loss of the condensation activity of the extract. In a control mitotic extract, sperm chromatin underwent a series of structural changes and was eventually transformed into a cluster of mitotic chromosomes (Fig. 2B) (12). In contrast, the chromatin was converted into a round structure in the Cdc2-depleted extract that was indistinguishable from the chromatin assembled in an interphase extract (Fig. 2B). When purified Cdc2–cyclin B (13) was added back into the depleted extract, the condensation activity was restored, accompanied by phosphorylation of the condensin subunits (Fig. 2A). The supercoiling activity of 13S condensin purified from the Cdc2-depleted extract was reduced relative to that from the control extract (Fig. 2C).

Figure 2

Control of supercoiling activity of 13Scondensin by Cdc2. (A) Effects of Cdc2 depletion. Mitotic extracts were immunodepleted with control immunoglobulin G (lane 1) or anti-Cdc2 (lanes 2 and 3) (11), and Cdc2–cyclin B purified from a Xenopus egg extracts (13) was added back to one portion of the depleted extract (lane 3). Amounts of Cdc2 protein (cdc2) and histone H1 kinase activity in the extracts were measured. Phosphorylation of condensin subunits was analyzed by immunoprecipitation from32P-labeled extracts ([32P]) or by immunoblotting with anti–XCAP-H (H-blot). (B) Condensation assay. Xenopus sperm chromatin was mixed with an interphase (I), mock-depleted mitotic (M), Cdc2-depleted mitotic (MΔcdc2), or Cdc2-reconstituted (MΔcdc2+cdc2) extract. After 3 hours at 22°C, chromatin was fixed and stained with 4′,6′-diamidino-2-phenylindole (DAPI) (12). (C) Supercoiling activity of 13S condensin purified from the mock-depleted (lanes 2 to 4) or Cdc2-depleted (lanes 6 to 8) extracts. Lanes 1 and 5, no protein. The molar ratios of protein to DNA were the same as in Fig. 1B.

These results suggest that 13S condensin is phosphorylated and activated either by Cdc2 itself or by kinases activated by Cdc2. Several consensus sites for phosphorylation by Cdc2 (14) exist in the sequences of XCAP-D2 (15, 16) and XCAP-H (4), and we tested whether purified Cdc2 phosphorylated these subunits in vitro. A purified Cdc2–cyclin B fraction phosphorylated the XCAP-D2 and XCAP-H subunits of 13S condensin isolated from an interphase extract (Fig. 3A) (17). This treatment converted the interphase 13S condensin into an active form that supported positive supercoiling of DNA (Fig. 3B). Two-dimensional tryptic phosphopeptide mapping (18) revealed three major peptides of XCAP-D2 phosphorylated by Cdc2–cyclin B that aligned with those labeled in mitotic extracts (Fig. 3C), suggesting that Cdc2 itself may phosphorylate XCAP-D2 in mitotic extracts. The maps of XCAP-H were more complex: At least 10 spots were detected in mitotic extracts, and five of them migrated with peptides phosphorylated by Cdc2–cyclin B. Thus, additional kinases are apparently required for full phosphorylation of XCAP-H. Nevertheless, after phosphorylation by Cdc2–cyclin B, the specific activity of 13S condensin from interphase extracts was comparable to that of 13Scondensin purified from mitotic extracts.

Figure 3

Phosphorylation and activation of 13S condensin by Cdc2. (A) Phosphorylation of XCAP-D2 and XCAP-H by Cdc2–cyclin B in vitro. 13S condensin was immunoprecipitated from32P-labeled interphase (lane 1) or mitotic (lane 2) extract (8). Alternatively, 13S condensin was purified from an interphase extract and then incubated with purified Cdc2–cyclin B in the presence of [γ-32P]ATP (lane 3) (17). The labeled proteins were analyzed by autoradiography (top) or by immunoblotting with anti–XCAP-H (bottom). (B) Supercoiling assay of 13S condensin phosphorylated by Cdc2–cyclin B. Interphase 13S condensin was phosphorylated by Cdc2–cyclin B (lanes 3, 6, and 9) or treated with buffer alone (lanes 2, 5, and 8) (17). Supercoiling assay was done in the presence of no topoisomerase (lanes 1 to 3) or type I topoisomerases from E. coli (lanes 4 to 6) or calf thymus (lanes 7 to 9). (C) Phosphopeptide mapping of XCAP-D2 (upper panels) and XCAP-H (lower panels) labeled by purified Cdc2–cyclin B (left) or in mitotic extracts (center) as described in (A) (18). (Right) spots overlapping under the two conditions are indicated by filled circles, and those unique to mitotic extracts are indicated by open circles.

To test whether Cdc2 consensus sites in the COOH-terminal region of XCAP-D2 are phosphorylated by Cdc2, we synthesized three phosphopeptides, each of which contained a single phosphothreonine, and prepared phospho-specific antibodies (19). The peptides were DP1 (EDDFQphosphoT1314PKPPA), DP2 (LSEAEphosphoT1348PKNPT), and DP3 (TPKNPphosphoT1353PIRRT) (16). Affinity-purified anti-DP1 recognized the mitotic form, but not the interphase form, of XCAP-D2, nor did it recognize the mitotic form that had been treated with λ-PPase (Fig. 4A). Antibody binding was blocked by the DP1 peptide, but not with an unphosphorylated peptide of the same sequence (DU1) or the other two phosphopeptides (Fig. 4B). Thus, anti-DP1 appears to recognize mitosis-specific phosphothreonine Thr1314 of XCAP-D2. Similarly, anti-DP2 and anti-DP3 recognized mitosis-specific phosphothreonines Thr1348 and Thr1353, respectively (Fig. 4, A and B). Immunodepletion of Cdc2 from a mitotic extract resulted in a loss of the three phosphoepitopes from XCAP-D2, and incubation of the interphase form of XCAP-D2 with purified Cdc2–cyclin B led to phosphorylation of these epitopes (Fig. 4A). Thus, the three sites clustered in the COOH-terminal domain are likely to be the physiological and direct targets of Cdc2. XCAP-D2 also acquired an MPM-2 epitope in a mitosis-specific and Cdc2-dependent manner [MPM-2 is a monoclonal antibody that recognizes mitosis-specific phosphoepitopes] (12, 20).

Figure 4

Identification of Cdc2 phosphorylation sites in the COOH tail of XCAP-D2. (A) Recognition of mitosis-specific phosphothreonines of XCAP-D2 by phosphopeptide antibodies. 13S condensin was purified from a mitotic extract (lanes 1 and 2), a Cdc2-depleted mitotic extract (lane 3), or an interphase extract (lanes 4 and 5), and then treated with λ-PPase (lane 2) or Cdc2–cyclin B (lane 5). The bands corresponding to XCAP-D2 were analyzed by immunoblotting with three phosphopeptide antibodies (anti-DP1, anti-DP2, and anti-DP3) (19), the phospho-specific mAb MPM-2 (20), or anti–XCAP-D2. (B) Specificity of phosphopeptide antibodies. 13S condensin was immunoprecipitated from a mitotic extract, and the band corresponding to XCAP-D2 was analyzed by immunoblotting with three phosphopeptide antibodies or anti–XCAP-D2. Peptide competitors (50 μdg/ml) were added to primary antibody solutions as indicated.

These results provide evidence for a direct functional link between the master mitotic kinase, Cdc2, and the key machinery of chromosome condensation. The supercoiling activity of 13S condensin may be a physiologically relevant activity that is essential for mitotic chromosome condensation.


Stay Connected to Science

Navigate This Article