Report

Mitotic and G2 Checkpoint Control: Regulation of 14-3-3 Protein Binding by Phosphorylation of Cdc25C on Serine-216

See allHide authors and affiliations

Science  05 Sep 1997:
Vol. 277, Issue 5331, pp. 1501-1505
DOI: 10.1126/science.277.5331.1501

Abstract

Human Cdc25C is a dual-specificity protein phosphatase that controls entry into mitosis by dephosphorylating the protein kinase Cdc2. Throughout interphase, but not in mitosis, Cdc25C was phosphorylated on serine-216 and bound to members of the highly conserved and ubiquitously expressed family of 14-3-3 proteins. A mutation preventing phosphorylation of serine-216 abrogated 14-3-3 binding. Conditional overexpression of this mutant perturbed mitotic timing and allowed cells to escape the G2 checkpoint arrest induced by either unreplicated DNA or radiation-induced damage. Chk1, a fission yeast kinase involved in the DNA damage checkpoint response, phosphorylated Cdc25C in vitro on serine-216. These results indicate that serine-216 phosphorylation and 14-3-3 binding negatively regulate Cdc25C and identify Cdc25C as a potential target of checkpoint control in human cells.

A key step in regulating the entry of eukaryotic cells into mitosis is the activation of the protein kinase Cdc2 by the protein phosphatase Cdc25C. A complete understanding of mitotic control requires elucidation of the mechanisms that regulate the interactions between Cdc2 and Cdc25C throughout the cell cycle. Furthermore, although tremendous progress has been made in recent years in identifying proteins that participate in checkpoint control, it is unclear how these proteins interface with core cell cycle regulators to inhibit cell cycle transitions (1).

The Ser216 residue is the primary site of phosphorylation of Cdc25C in asynchronously growing cells (2). To determine if phosphorylation of Ser216 regulates Cdc25C function, we generated HeLa cell lines that allow conditional expression of either wild-type Cdc25C or a mutant of Cdc25C containing alanine at position 216 (S216A). In these cells, expression of Cdc25C and Cdc25(S216A) is under the control of a hybrid protein consisting of a bacterial tetracycline repressor and VP16 activator protein (3). Expression can be induced upon removal of tetracycline from the medium (Fig.1A). Two electrophoretic forms were evident in the case of induced Cdc25C, a major form (species b) and a minor form (species a). A single electrophoretic form migrating in the position of species a was evident for induced Cdc25(S216A). Phosphatase treatment converted the electrophoretic mobility of species b to that of species a, demonstrating that phosphorylation of Ser216 is responsible for the shift in mobility of species b (4). The electrophoretic mobility of Cdc25C in SDS gels could therefore be used to monitor amounts of Ser216phosphorylation in vivo. Phosphopeptide mapping experiments (5) revealed one major and several minor phosphopeptides for induced Cdc25C (Fig. 1B). The major phosphopeptide contains Ser216 (2) and was absent in maps of induced Cdc25(S216A).

Figure 1

Phosphorylation of Cdc25C on Ser216 in asynchronously growing HeLa cells. (A) Expression of Cdc25C and Cdc25(S216A) was induced by removal of tetracycline (Tet) from the media (3). At the indicated times (hours), cells were collected, and Cdc25C was visualized by immunoblotting. Species a and b refer to different electrophoretic forms of Cdc25C. (B) Induced cells were incubated with32P-labeled inorganic phosphate. Radiolabeled Cdc25C (left panel) and Cdc25(S216A) (right panel) were isolated by immunoprecipitation and subjected to two-dimensional tryptic phosphopeptide mapping. Arrows depict the origin.

To determine whether phosphorylation of Cdc25C at Ser216 was regulated by the cell cycle, we elutriated Jurkat cells and analyzed fractions for endogenous Cdc25C mobility by immunoblotting (6). Most Jurkat cell Cdc25C was phosphorylated on Ser216 in asynchronously growing cells and throughout the G1 and S phases of the cell cycle (Fig. 2A). The G2-M population of Jurkat cells contained Cdc25C phosphorylated on Ser216 but also contained the mitotic form of Cdc25C. We used a double-thymidine block and release protocol to monitor Ser216 phosphorylation in synchronized populations of HeLa cells (7). Amounts of Cdc25C phosphorylated at Ser216 remained constant throughout the first 9 hours after release, corresponding to passage through the S and G2 phases of the cell cycle (Fig.2B). Mitotic cells collected between 10 and 12 hours after release were enriched in the mitotic form of Cdc25C and had lower amounts of the Ser216-phosphorylated form of Cdc25C. Phosphopeptide maps of Cdc25C and Cdc25(S216A) were generated from induced HeLa cells incubated with 32P during mitosis (Fig.2C) (5). Cdc25C and Cdc25(S216A) yielded identical phosphotryptic maps, and the Ser216-containing phosphopeptide was not detected (this peptide migrates between phosphopeptides 7 and 8). Thus, Cdc25C is phosphorylated on Ser216 throughout interphase but not during mitosis.

Figure 2

Cell cycle regulation of Ser216 phosphorylation. (A) Asynchronous and elutriated populations of Jurkat cells were lysed, resolved by SDS-PAGE, and immunoblotted for Cdc25C. Lane 1, asynchronous population of Jurkat cells; lanes 2, 3, and 4, Jurkat cells enriched in G1, S, or G2-M, respectively. Species b is Cdc25C phosphorylated on Ser216 and species c is the mitotic form of Cdc25C. (B) HeLa cells were synchronized at G1-S by a double-thymidine block (time 0, lane 1) and then allowed to proceed through S (4 hours, lane 2), G2 (8 and 9 hours, lanes 3 and 4), and into mitosis. Mitotic cells were collected by shake-off between 10 and 12 hours after release (lane 5). Synchronized cells were lysed, resolved by SDS-PAGE, and immunoblotted for Cdc25C. (C) Induced cells were arrested in mitosis with nocodazole and incubated with 32P-labeled inorganic phosphate. Mitotic forms of Cdc25C (left panel) and Cdc25(S216A) (right panel) were isolated by immunoprecipitation and subjected to two-dimensional tryptic phosphopeptide mapping. Arrows depict the origin.

We used chromosome spreading to assess the role of Ser216phosphorylation in regulating mitotic entry (8,9). Normal mitotic cells display intact chromosomes upon spreading, whereas chromosomes from cells that enter mitosis from S phase fragment upon spreading. Less than 0.1% of the mitotic nuclei derived from cells induced to express Cdc25C showed abnormal chromosome spreads, whereas 4.4 ± 0.7% of the mitotic nuclei derived from cells induced to express Cdc25(S216A) showed abnormalities (Table1).

Table 1

Effects of Cdc25(S216A) overexpression on mitotic and replication checkpoint control. Uninduced cells or cells induced to express either Cdc25C or Cdc25(S216A) were incubated for 8 hours in the presence of nocodazole (for examination of mitotic control) or were incubated for 8 hours in the presence of hydroxyurea followed by an additional 8 hours in the presence of hydroxyurea and nocodazole (for examination of replication checkpoint control). Cells were processed for chromosome spreading as described (8). Mitotic nuclei were identified from a population of 2000 nuclei each in three independent experiments. Data are presented as mean ± SEM.

View this table:

Cells induced to express Cdc25C and Cdc25(S216A) were also analyzed for DNA replication and DNA damage checkpoint responses (9). A DNA replication checkpoint response was induced by incubation of the cells in the presence of hydroxyurea. Most cells induced to express Cdc25C were arrested in S phase as indicated by the low numbers of mitotic nuclei (Table 1). In contrast, five times more mitotic nuclei were counted in cells induced to express Cdc25(S216A) (Table 1). Although the number of normal nuclei were similar for uninduced and induced cells, cells induced to express Cdc25(S216A) had significantly more abnormal nuclei (Table 1). In addition, Cdc25(S216A) cells had detectable amounts of the mitotic form of Cdc25C and larger amounts of cyclin B1–associated histone H1 kinase activity (4). Finally, to generate a DNA damage checkpoint response, cells were gamma irradiated and monitored for their ability to delay in G2(9). In the absence of irradiation, the S phase cells from uninduced and Cdc25(S216A)-induced populations took between 8 and 12 hours to reach G1 (Fig. 3, A and B). At 24 hours after irradiation, only 9% of the S phase cells from uninduced populations had cycled to G1, indicating a 12- to 16-hour radiation-induced delay (Fig. 3C). Induced expression of Cdc25C resulted in a partial loss in the G2 delay, which was exaggerated in the case of Cdc25(S216A) expression (Fig. 3, D, G, and H). By 12 hours, 9% of the Cdc25(S216A)-expressing cells were already in G1 and 16% were in G1 by 16 hours. After 24 hours, the G1-S fractions represented 60% of the population. Cycling and arrest of BrdU-positive cells demonstrated that the observed results are due to a G2 delay rather than to cells not releasing from G1 (Fig. 3, A, B, C, and D).

Figure 3

Disruption of DNA damage checkpoint in Cdc25(S216A) cells. Uninduced cells or cells induced to express Cdc25C or Cdc25(S216A) were labeled with BrdU and then treated with 0 or 6 Gy of gamma irradiation. Cells were harvested at the indicated times and stained with PI and for BrdU as described (9). DNA content profiles for BrdU-positive cells are shown for uninduced and Cdc25(S216A) induced cells in (A) to (D). Total DNA content profiles and quantitation of G1, S, and G2-M irradiated cells are shown in (E) to (H).

The sequence bordering and inclusive of Ser216 in Cdc25C (RSPS216MP) contains a potential recognition motif for binding of 14-3-3 proteins: RSXSXP where P is proline, R is arginine, X is any amino acid, and the underlined serine is phosphorylated (10). The 14-3-3 proteins belong to a highly conserved multigene family of small acidic proteins that associate with cell cycle and cell death regulators, oncogenes, and signaling molecules (11, 12). There are at least seven mammalian 14-3-3 isoforms ranging in size from 30 to 35 kD. Immunoblotting experiments were performed with a 14-3-3 antibody that recognizes several of the mammalian isoforms to determine whether a complex between Cdc25C and 14-3-3 proteins could be detected. We detected 14-3-3 proteins in immunoprecipitates of Cdc25C but not Cdc25(S216A) (Fig. 4A). Given that the mitotic form of Cdc25C was not detectably phosphorylated on Ser216 (Fig. 2C), we determined whether the binding of Cdc25C to 14-3-3 was lost during mitosis. HeLa cells induced to express Cdc25C were arrested in mitosis with nocodazole (3), and Cdc25C immunoprecipitates were monitored for the presence of 14-3-3 by immunoblotting. No 14-3-3 was immunoprecipitated with the mitotic form of Cdc25C (Fig. 4A). We used a double-thymidine block and release protocol to monitor 14-3-3 binding at other phases of the cell cycle (6). Binding of 14-3-3 was detected during the G1, S, and G2 phases of the cell cycle (Fig.4B). Binding of 14-3-3 was reduced in fractions enriched for M phase cells.

Figure 4

Association of Cdc25C with 14-3-3 in a phosphorylation- and cell cycle–dependent manner. (A) Cells induced to express Cdc25C and Cdc25(S216A) were untreated (lanes 1 and 3) or were incubated in the presence of nocodazole (lane 2). Immunoprecipitates were prepared from 3 mg of total cellular protein by using 9E10 Myc-agarose and analyzed for Cdc25C (upper panel) and 14-3-3 (lower panel) by immunoblotting. AS, asynchronous cells; M, mitotic cells. (B) Cells induced to express Cdc25C were synchronized at G1-S by a double-thymidine block (time 0, lane 1) and then allowed to proceed through S (4 hours, lane 2), G2 (9 hours, lane 3), M (10 to 12 hours, lane 4), and G1 (13 hours, lane 5). Cells were harvested at the indicated times, and immunoprecipitates of Cdc25C were prepared from 2 mg of total cellular protein. Cdc25C (upper panel) and 14-3-3 (lower panel) were visualized by immunoblotting. (C) Coprecipitates of GST-Cdc25C and 14-3-3 from overproducing insect cells were incubated in vehicle (lanes 1 and 2) or vehicle containing unphosphorylated (lanes 3 and 4) or Ser216-phosphorylated (lanes 5 and 6) peptide at concentrations of 10 and 100 μM. Reactions were resolved by SDS-PAGE and immunoblotted for Cdc25C and 14-3-3.

Mutation of Ser216 abrogated the binding of 14-3-3 to Cdc25C, demonstrating that Ser216 is essential for the interaction. To determine whether phosphorylation of Ser216 was important for the interaction, we performed competition experiments using peptides consisting of amino acids 210 to 225 of Cdc25C that were either phosphorylated on Ser216 or unphosphorylated (13). Complexes consisting of 14-3-3 and Cdc25C-GST fusion protein purified from insect cells were disrupted by incubation with excess peptide containing phosphorylated Ser216 but not by unphosphorylated peptide (Fig. 4C). These results demonstrate that the phosphorylation of Ser216is required for 14-3-3 binding to Cdc25C.

Mitotic hyperphosphorylation of Cdc25C on NH2-terminal serine and threonine residues increases its intrinsic phosphatase activity (1). In contrast, phosphorylation of Cdc25C on Ser216 throughout interphase appears to negatively regulate Cdc25C. Our results suggest that the negative effects of Ser216phosphorylation may be mediated by 14-3-3 binding. We have demonstrated that 14-3-3 is bound to Cdc25C during phases of the cell cycle when Cdc25C is phosphorylated on Ser216and functionally inactive and is released in mitosis when Cdc25C is maximally active and not phosphorylated on Ser216. We propose that Ser216phosphorylation and 14-3-3 binding sequester Cdc25C from functionally interacting with Cdc2 in vivo, because the phosphatase activity of Cdc25C was not detectably altered in response to either Ser216 phosphorylation or 14-3-3 binding (14).

The fission yeast homologs of 14-3-3, Rad24 and Rad25, have been shown to play a role in mitotic and radiation checkpoint control (15, 16). Loss of either gene causes early entry into mitosis and partial loss of the radiation checkpoint, similar to the phenotype reported here for cells expressing the Cdc25C mutant (S216A). An inability to inhibit fission yeast Cdc25 activity could account for the observed rad24 + andrad25 + mutant phenotypes. The Chk1 protein kinase is another essential component of the DNA damage checkpoint in fission yeast (17-19). Cells that lackchk1 +are viable but fail to delay mitotic entry in response to damaged DNA and subsequently die. We tested whether Chk1 from Schizosaccharomyces pombe could phosphorylate Cdc25C in vitro (20). Chk1 phosphorylated both full-length Cdc25C and a GST fusion protein consisting of amino acids 200 to 256 of Cdc25C (Fig. 5A). Phosphoamino acid analysis revealed phosphoserine (21), and trypsin digestion of Cdc25C followed by high-pressure liquid chromatography (HPLC) analysis gave rise to a single phosphopeptide that eluted in fraction 57 (Fig. 5C). Sequencing of this tryptic phosphopeptide before and after digestion with proline-specific endopeptidase identified Ser216 as the site of phosphorylation. Human Chk1 also phosphorylated Cdc25C on Ser216, demonstrating the conservation of this regulatory pathway (22). The ability of both fission yeast and human Chk1 to phosphorylate Cdc25C on Ser216 implicates Chk1 as possibly regulating the interactions between 14-3-3 and Cdc25C during a DNA damage checkpoint response.

Figure 5

Phosphorylation of Cdc25C on Ser216 by S. pombe Chk1 in vitro. (A) GST-Chk1 bound to GSH-agarose was tested for its ability to phosphorylate soluble His6Cdc25C (lanes 1 and 2) and a GST fusion protein consisting of amino acids 200 to 256 of Cdc25C (lane 3). Reactions were separated into pelleted (lane 1) and supernatant (lane 2) fractions by centrifugation or were analyzed directly (lane 3). Proteins were resolved on a 7% gel and visualized by autoradiography. (B) Amino acids inclusive of and surrounding Ser216 showing NH2-terminal trypsin and proline endopeptidase cleavage sites. (C) Radiolabeled His6-Cdc25C was digested with trypsin, and the tryptic peptides were resolved by reverse-phase HPLC. Column fractions were collected and monitored for the presence of radioactivity. (D) Manual Edman degradation of tryptic phosphopeptide present in fraction 57 (right panel). The dotted lines indicate radioactivity remaining bound to the sequencing membrane at the end of each cycle, and bars represent radioactivity released from the membrane.

  • * To whom correspondence should be addressed. E-mail: hpiwnica{at}cellbio.wustl.edu

REFERENCES AND NOTES

View Abstract

Navigate This Article