Conservation of the Chk1 Checkpoint Pathway in Mammals: Linkage of DNA Damage to Cdk Regulation Through Cdc25

See allHide authors and affiliations

Science  05 Sep 1997:
Vol. 277, Issue 5331, pp. 1497-1501
DOI: 10.1126/science.277.5331.1497


In response to DNA damage, mammalian cells prevent cell cycle progression through the control of critical cell cycle regulators. A human gene was identified that encodes the protein Chk1, a homolog of the Schizosaccharomyces pombe Chk1 protein kinase, which is required for the DNA damage checkpoint. Human Chk1 protein was modified in response to DNA damage. In vitro Chk1 bound to and phosphorylated the dual-specificity protein phosphatases Cdc25A, Cdc25B, and Cdc25C, which control cell cycle transitions by dephosphorylating cyclin-dependent kinases. Chk1 phosphorylates Cdc25C on serine-216. As shown in an accompanying paper by Peng et al. in this issue, serine-216 phosphorylation creates a binding site for 14-3-3 protein and inhibits function of the phosphatase. These results suggest a model whereby in response to DNA damage, Chk1 phosphorylates and inhibits Cdc25C, thus preventing activation of the Cdc2–cyclin B complex and mitotic entry.

Cell cycle checkpoints are regulatory pathways that control the order and timing of cell cycle transitions and ensure that critical events such as DNA replication and chromosome segregation are completed with high fidelity. In response to DNA damage, cells activate a checkpoint pathway that arrests the cell cycle to provide time for repair and induces the transcription of genes that facilitate repair. In yeast, this checkpoint pathway consists of several protein kinases including phosphoinositide (PI)–kinase homologs hATM, scMec1, and spRad3 and protein kinases scDun1, scRad53, and spChk1 (1) (the prefixes h, sc, and sp refer to Homo sapiens, Saccharomyces cerevisiae, andSchizosaccharomyces pombe, respectively). In mammals, this pathway results in the activation of p53, which induces transcription of the cyclin-dependent kinase inhibitor p21CIP1, resulting in arrest in the G1 phase of the cell cycle (2).

To address the conservation of checkpoint function we searched for human homologs of yeast checkpoint genes. We used a degenerate polymerase chain reaction (PCR) strategy and identified a human gene very similar to the gene encoding Chk1 in S.pombe (Fig. 1) (3). With human CHK1 cDNA as a probe, we isolated the gene encoding Chk1 from mouse (mChk1). The sequence of the longest human cDNA (1891 base pairs) predicted a translation product of 476 amino acids with a molecular size of 54 kD (Fig. 1A). No in-frame stop codon was found upstream of the first methionine, which is within the Kozak consensus sequence (4) and is likely to be the initiation codon because its encoded protein is the same size as that observed in cells (see below). The human CHK1 gene is related to aCaenorhabditis elegans gene in the database and aDrosophila melanogaster gene grp, which has a role in cell cycle control and development (5) (Fig. 1B). The predicted hChk1 protein is 29% identical and 44% similar to spChk1, 40% identical and 56% similar to the ceChk1 (ce referes toC. elegans), and 44% identical and 56% similar to dmChk1 (dm refers to D. melanogaster). Sequence analysis revealed several COOH-terminal domains that are highly conserved in the Chk1 family of kinases (Fig. 1B).

Figure 1

Isolation of the human and mouseCHK1 genes. (A) Domain structure of the predicted human Chk1 (hChk1) protein. The black boxes indicate regions of highest conservation. The GenBank accession number for hChk1 isAF016582, and for mChk1, AF016583. (B) Alignment of Chk1 homologs. Amino acid identities are shown as black boxes. Conservative changes are shown as shaded boxes. Hs is Homo sapiens, Sp is S. pombe, Ce is C. elegans, and Dm isD. melanogaster. The database DNA sequence for ceChk1 has a likely frame shift in the COOH-terminus. Singleletter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.

The chromosomal location of CHK1 was mapped to 11q24 by fluorescence in situ hybridization (Fig.2A). This is adjacent to the gene encoding ATM at 11q23. Loss of heterozygosity at this region has been associated with a number of cancers including those of the breast, lung, and ovaries (6). Northern (RNA) blot analysis revealed ubiquitous expression of hChk1, with large amounts in human thymus, testis, small intestine, and colon (Fig. 2B). In adult mice, mChk1 was detected in all tissues examined and in large amounts in the testis, spleen, and lung (Fig. 2B). Mouse embryos from embryonic day 15.5 also revealed ubiquitous expression, with large amounts detected in the brain, liver, kidney, pancreas, intestines, thymus, and lung (7). Testis, spleen, and thymus also express large amounts of ATM (8).

Figure 2

Localization of CHK1 on chromosome 11q24. (A) In situ hybridization was performed on mitotic chromosomes with fluorescently labeled human CHK1 DNA. Arrows indicate CHK1 localization at 11q24. (B) Northern analysis of human and mouse Chk1. Blots containing the polyadenylated RNA (2 μg per lane) from the indicated tissues were probed with human or mouse CHK1 cDNAs.

Affinity-purified antibodies to hChk1 protein made in baculovirus (anti-FL) (9) or to its COOH-terminal 15 amino acids (anti-PEP) recognized a 54-kD protein (Fig.3A) that comigrates with hChk1 expressed in baculovirus (7). The anti-PEP but not anti-FL signal was competed by addition of excess peptide, indicating that the two sera recognize different hChk1 epitopes, further confirming identity of the 54-kD band as endogenous hChk1. A 70-kD protein was also specifically recognized by anti-PEP. When mCHK1 was expressed from the cytomegalovirus (CMV) promoter in baby hamster kidney (BHK) cells, we detected a 54-kD nuclear protein only in transfected cells using antibodies to the COOH-terminal peptide of mChk1. This exogenous mChk1 comigrates with endogenous mChk1 from mouse lung tissue (7).

Figure 3

Modification of the 54-kD nuclear hChk1 protein in response to DNA damage. (A) Protein from the indicated cell lines was fractionated by SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotted with either affinity-purified antibodies to a COOH-terminal peptide (anti-PEP) or full-length hChk1 (anti-FL). HeLa extracts were immunoblotted with anti-PEP or anti-FL in the presence or absence of the 15–amino acid peptide used to produce the anti-PEP. (B) Modification of hChk1 in response to DNA damage. For the top panel HeLa cells were synchronized with 2 mM thymidine and treated without (–) or with (+) 10 Gy of ionizing radiation 1 hour after release from the thymidine block. Cells were collected in G2-M, and extracts were fractionated by 10% SDS-PAGE and immunoblotted with anti-PEP. For the bottom panel Jurkat cells were treated (+IR) or not treated (–IR) with 10 Gy of ionizing radiation and incubated for 2 hours. Extracts from these cells were resolved in the first dimension by using isoelectric focusing (IEF) with pH 3 to 10 ampholytes, in the second dimension on a 10% SDS-PAGE, and immunoblotted with anti-PEP. (C to F) Punctate nuclear localization of hChk1. Human fibroblasts were fixed, stained with 4′,6′-diamidino-2-phenylindole (DAPI) to detect DNA, and probed with affinity-purified anti-PEP, biotinylated antibody to rabbit immunoglobulin G, and Texas Red streptavidin to reveal subcellular localization. (C) and (E), DAPI-stained nuclei; (E) and (F), hChk1 protein. Original magnifications: (C) and (D), ×40; (E) and (F), ×150.

To determine whether hChk1 is modified in response to DNA damage like spChk1, we examined hChk1 protein in extracts from cells treated with ionizing radiation. hChk1 from extracts from damaged cells showed a minor but reproducible reduction in mobility compared with hChk1 from untreated cells (Fig. 3B). The change in mobility observed in response to DNA damage for spChk1 was also slight (10). This modification was confirmed by two-dimensional gel analysis, which demonstrated the generation of a more negatively charged hChk1 species 2 hours after gamma irradiation (Fig. 3B). These results indicate that hChk1 may participate in transduction of the DNA damage signal like spChk1. Indirect immunofluorescence revealed that hChk1 is localized to the nucleus in a punctate staining pattern (Fig.3C), similar to that observed for ATM (8). mChk1 expressed in BHK cells confirmed the nuclear localization (7).

To test for the ability of hChk1 to regulate the cell cycle, we transfected hChk1 or hChk1(D130A), a catalytically inactive mutant, under the control of the CMV promoter or the CMV vector alone into HeLa cells treated with and without 6 Gy of ionizing radiation. We did not detect perturbation of the cell cycle by either kinase relative to vector alone, suggesting that overproduction alone was insufficient to deregulate the system (7).

Tyrosine phosphorylation of Cdc2 has been implicated in cell cycle arrest in response to DNA damage and replication blocks in both S. pombe (11, 12) and humans (13). In S. pombe, Cdc2 mutants that cannot be phosphorylated on tyrosine are unable to arrest the cell cycle in response to blockade of DNA replication. Although it was originally thought that the DNA damage checkpoint did not operate through tyrosine phosphorylation, recent experiments have shown that tyrosine phosphorylation is required forS. pombe cells to arrest in response to DNA damage (12, 14). Although it is now clear that tyrosine phosphorylation is required for proper checkpoint control, the experiments implicating tyrosine phosphorylation in this pathway do not distinguish between a regulatory role in which tyrosine phosphorylation rates are manipulated by the checkpoint pathways, or a passive role in which tyrosine phosphorylation is required to allow cell cycle arrest but is not the actual target of the checkpoint pathway (1, 15).

To address this issue, we analyzed the ability of hChk1 to phosphorylate key regulators of Cdk tyrosine phosphorylation, the Cdc25 dual-specificity phosphatases hCdc25A, hCdc25B, and hCdc25C. These regulators were singled out for several reasons. First, overproduction of hCdk4 mutants in which the inhibitory tyrosine is changed to phenylalanine abrogates G1 arrest in response to ultraviolet (UV) light (16). Second, the UV sensitivity ofchk1 mutants in S. pombe is suppressed by inactivating cdc25 with a temperaturesensitive mutation (10). Finally, in S. pombe wee1 mik1 mutants, DNA damage still causes a partial cell cycle delay that could be due to regulation of spCdc25 activity (12). GST-hChk1 and GST-hChk1(D130A) were introduced into baculovirus, purified from baculovirus-infected insect cells, and incubated with GST-hCdc25A, GST-hCdc25B, and GST-hCdc25C (9, 17). GST-hChk1 phosphorylated all three hCdc25 proteins but not GST alone (Fig. 4A). Although GST-hCdc25C comigrated with GST-hChk1, which autophosphorylates, increased phosphorylation was observed at that position relative to phosphorylation in the presence of kinase alone, and phosphorylation of a GST-hCdc25C breakdown product was visible. In separate experiments with a His6-tagged hChk1 derivative, there was phosphorylation of GST-hCdc25C (Fig.4B). A catalytically inactive mutant failed to phosphorylate itself or any of the hCdc25 proteins (Fig. 4A).

Figure 4

hChk1 binds to and phosphorylates hCdc25A, hCdc25B, and hCdc25C. (A) GST-hChk1 (W) and GST-hChk1(D130A) (k) were purified from baculovirus and incubated with either GST, GST-hCdc25A, GST-hCdc25B, GST-hCdc25C, or GST-Cdc25C(200–256) and [γ32P]ATP. Proteins were resolved by SDS-PAGE (10%) and visualized by autoradiography (for the kinase assay, top) or Coomassie blue staining (bottom). Less GST-Cdc25B was loaded than the other substrates. GST-hChk1 did not phosphorylate GST alone. (B) For the left panels, GST-hChk1 (WT) from baculovirus was incubated with either GST-hCdc25C(200–256) or GST-hCdc25C(200–256)(S216A), and [γ32P]ATP. For the right panels, hChk1-His6 purified from baculovirus was incubated with either GST-hCdc25C (WT) (lane 5) or GST-hCdc25C(S216A) and [γ32P]ATP. Proteins were resolved and visualized for top and bottom panels as in (A). (C) Extracts prepared from insect cells infected with hChk1-His6 expressing baculovirus were incubated with bacterially expressed GST, GST-hCdc25A, GST-hCdc25B, and GST-hCdc25C bound to GSH beads, precipitated, fractionated on SDS-PAGE, and probed with affinity-purified anti-PEP to hChk1. (D) Amino acids inclusive of and surrounding Ser216 showing NH2-terminal trypsin and proline endopeptidase cleavage sites. His6-Cdc25C radiolabeled by GST-hChk1 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 (bottom, left panel). Manual Edman degradation of tryptic phosphopeptide present in fraction 57 (bottom, 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.

Protein kinases often form complexes with their substrates. To see if this was the case for hChk1 and the Cdc25 proteins, GST-hCdc25 proteins on glutathione beads were incubated together with baculovirus extracts expressing His6-tagged hChk1 and precipitated. GST-hCdc25A, GST-hCdc25B, and GST-hCdc25C each specifically bound hChk1 whereas GST alone did not (Fig. 4C). Furthermore, two other GST fusion proteins, GST-Dun1 and GST-Skp1, failed to bind hChk1 (18). These results indicate that Cdc25 can form complexes with hChk1.

To establish the significance of the Cdc25 phosphorylation, we mapped the site of hChk1 phosphorylation on Cdc25C. The Ser216 residue is the main site of phosphorylation of hCdc25C in vivo (19). hChk1 phosphorylated a 56–amino acid region of the hCdc25C protein fused to GST (19) but not GST alone (Fig. 4A). This 56–amino acid motif contains four possible sites of phosphorylation. Peptide analysis of proteolytic fragments of full-length His6-hCdc25C phosphorylated by GST-hChk1 revealed a single phosphorylated tryptic peptide by HPLC. Edman degradation of this peptide indicated release of radioactivity in the third cycle (Fig. 4D). Further degradation of this tryptic fragment with proline endopeptidase resulted in a peptide that released radioactivity in the first cycle (17). The Ser216 residue is the only site on hCdc25C consistent with this phosphorylation pattern (Fig. 4D). To confirm this, we constructed the Cdc25C S216A mutation in GST-Cdc25C and Cdc25C(200–256). Both were poor substrates for hChk1, confirming Ser216 as the site of phosphorylation (Fig.4C). Serine-216 is also phosphorylated by spChk1, demonstrating phylogenetic conservation of this regulatory relation (20).

We have shown that the Chk1 kinase family is conserved throughout eukaryotic evolution and that hChk1, like its S. pombecounterpart, is modified in response to DNA damage. This, together with the fact that ATM-related kinases are conserved members of checkpoint pathways and act upstream of chk1 in S. pombe, suggests that this entire checkpoint pathway may be conserved in all eukaryotes. hChk1 directly phosphorylates a regulator of Cdc2 tyrosine phosphorylation, hCdc25C, on a physiologically significant residue, Ser216. Support for this comes from the work of Peng et al. (20) who have shown that the same site, which is the major site of Cdc25C phosphorylation during interphase, binds 14-3-3 proteins when phosphorylated and acts in an inhibitory fashion on hCdc25C. Overexpression of the hCdc25C S216A protein reduces the ability of cells to arrest in G2 in response to DNA damage as observed previously for the Cdc2AF mutants (12). The overexpression studies alone do not prove that the DNA damage checkpoint pathway operates through tyrosine phosphorylation, because hyperactive Cdc2 may be able to bypass checkpoint control. However, in combination with the fact that this inhibitory serine is directly phosphorylated by the DNA damage–responsive checkpoint kinase hChk1, these results strongly imply that DNA damage regulates the G2-to-mitosis transition through control of Cdc2 tyrosine phosphorylation. These results suggest a model whereby in response to DNA damage, hChk1 phosphorylates hCdc25C on Ser216, which leads to binding of 14-3-3 protein and inhibition of Cdc25C's ability to dephosphorylate and activate Cdc2, a model that will require genetic verification. This model does not preclude a role for other cell cycle regulators such as Wee1 in the damage response (14). Furthermore, the fact that hChk1 phosphorylated hCdc25A and hCdc25B and that Ser216 is conserved among these Cdc25 proteins (19) suggests that hChk1 may regulate other DNA damage checkpoints, such as those controlling the G1-to-S phase transition, through a similar mechanism.

  • * To whom correspondence should be addressed. E-mail: selledge{at}


View Abstract

Navigate This Article