Introduction

Chk2, the mammalian homolog of the yeast Rad53 and Cds1 genes, encodes a nuclear serine/threonine kinase that plays a crucial role in the DNA damage response and helps guard the integrity of the genome by regulating cell-cycle checkpoints, DNA repair and apoptosis.1, 2, 3 Following DNA damage, Chk2 is phosphorylated and activated by ataxia-telangiectasia-mutated (ATM)-dependent and -independent mechanisms,4, 5, 6, 7, 8 and exerts diverse biological effects via phosphorylation of its downstream effector molecules, including p53, Cdc25, BRCA1, PML and E2F-1.9, 10, 11, 12, 13, 14 Importantly, germ-line mutations within the Chk2 gene have been reported in a subset of families with Li-Fraumeni syndrome (LFS), but who have a normal, wild-type p53,15, 16, 17 suggesting that Chk2 may function as a tumor suppressor. While mutations in the Chk2 gene have been implicated in occasional sporadic tumors,3, 18, 19, 20, 21, 22 somatic Chk2 mutations are rare in particular in lymphoid malignancies.23, 24, 25

Lymphoid malignancies are a heterogeneous group of neoplasms characterized by distinct clinical, immunophenotypical, morphological, cytological and genetic features. Alterations in tumor suppressors that are involved in the DNA damage response have been reported to be frequently involved in the pathogenesis of lymphoid malignancies. In particular, the incidence of p53 mutations is apparently high in lymphoid malignancies, and is associated with the aggressive clinical development of these neoplasms.26, 27 It has been reported that AT patients have a 100-fold increased risk for the development of lymphoid malignancies.28 Furthermore, mutations in the ATM gene have been associated with the development of chromosomal instability in a subset of lymphoid malignancies as a result of impairment in the DNA damage response29, 30 Since Chk2 acts downstream of ATM and upstream of p53, it is anticipated that mutations in Chk2 may also be involved in the onset or pathogenesis of lymphoid neoplasms.

In those infrequent cases where somatic Chk2 mutations do occur in sporadic tumors, including lymphoid malignancies, it has been found that nonsense or missense mutations often result in downregulated Chk2 expression or impaired kinase activity.20, 31 It has also been recently reported that alterations in the post-transcriptional regulation of Chk2 expression occur in a subset of aggressive non-Hodgkin's lymphomas (HLs).32 In these cases, Chk2 protein is poorly expressed or absent, yet Chk2 mRNA is apparently expressed normally compared to other types of lymphomas, and no apparent mutation or altered methylation is found in the Chk2 gene. On the other hand, in HL, malignant mononuclear Hodgkin and multinucleated Reed-Sternberg (H/RS) cells are occasionally found to harbor genetic and/or epigenetic alterations (e.g. mutation, splicing variants and DNA methylation) in genes encoding several cell-cycle regulators, including p16INK4a, p14ARF and IκBα.33, 34, 35, 36 While p53 mutations are rather rare in H/RS cells,37 the expression of genes or proteins involved in the DNA damage response in these cells has not been thoroughly examined.

Accumulating evidence demonstrates the importance of post-translational modification of histone proteins (i.e. acetylation, methylation and phosphorylation), in addition to DNA methylation, as epigenetic mechanisms involved in the organization of chromosomal domains and gene regulation.38, 39, 40, 41 It is now generally appreciated that hyperacetylated histones H3 and H4 and methylated H3-lysine 4 are associated with activated genomic regions at both the local and global levels, while hypoacetylation of histones H3, H4 and methylation on H3-lysine 9 results in gene repression and silencing. Interestingly, it has been shown that there exists an interplay between DNA methylation and histone modification (methylation and acetylation). In fact, methylation of histone H3-lysine 9, associated with aberrant gene silencing of the p14ARF/p16INK4a locus in tumor cells, is rapidly reversed by treatment with a DNA methyltransferase inhibitor, 5-Aza-2′-deoxycytidine (5Az-dC).42

In this study we first examined the expression levels of Chk2, p53 and Wip1 in nine cell lines from various lymphoid malignancies, including three HL lines. Chk2 mRNA and protein expression was dramatically lower in all three HL cell lines compared to the other cell lines, yet there was no apparent mutation in the Chk2 gene. Therefore, we tested the possibility that epigenetic mechanisms are involved in this aberrant expression of the Chk2 gene in these cells using the histone deacetylase (HDAC) inhibitors trichostatin A (TsA) and sodium butyrate (SB). Histone acetylation was assayed by a chromatin-immunoprecipitation (Chip) method. Our results indicate that decreased expression of the Chk2 gene in HL cells is due, at least in part, to hypoacetylation of histones H3 and H4. We further investigated the possible hypermethylation of the Chk2 gene in HL cells by DNA methylation assays, but found that the global methylation patterns in the Chk2 gene and the methylation status of a likely methylation site, a CpG-rich region upstream of the Chk2 gene, were unaffected in HL cells. However, treatment of HL cells with 5Az-dC resulted in the hypomethylation of histone H3-lysine 9 and hyperacetylation of histones H3 and H4, leading the restoration of normal Chk2 expression. These findings indicate that DNA methylation of Chk2 at an as-yet unidentified site(s) is also involved in the downregulation of Chk2 expression, and that there exists an interplay between DNA methylation and histone modification in regulating the expression of Chk2 in HL cells. We also discuss the possible relationship between downregulation of Chk2 expression and the oncogenic properties of HL cells.

Results

Expression of Chk2 mRNA and protein in cell lines from HLs and other lymphoid malignancies

We first performed Northern blot analysis to examine the expression levels of Chk2, p53 and Wip1 mRNAs in nine cell lines from various lymphoid malignancies; HDLM2 and L428 (HLs, nodular sclerosis), KM-H2 (HL, mixed cellularity), KM3 (pre-B lymphocytic leukemia), RL (non-HL), FL318 and FL518 (follicular lymphomas), Black93 (Burkitt's lymphoma), and Jurkat (T cell leukemia) (see Materials and Methods). Although the expression patterns of the Chk2, p53 and Wip1 mRNAs were different among the various cell lines examined, each transcript was detected at the predicted size, with the exception of p53 mRNA in HDLM2 and Jurkat cells (Figure 1a). Interestingly, all the three HL cells, HDLM2, L428 and KM-H2 cells, exhibited drastically decreased expression of Chk2 mRNA compared to the other six cell lines (Figure 1a). The Chk2 genes from HL cells (L428 and KM-H2 cells) and from KM3 and RL cells were not found to possess any apparent mutations within their exons by sequence analyses (data not shown). These results suggest that the Chk2 mRNA is altered at the transcriptional or post-transcriptional level in HL cells.

Figure 1
figure 1

Expression levels of Chk2 mRNA and protein in cell lines from various malignancies. (a) Expression levels of Chk2, p53 and Wip1 mRNAs in the indicated cell lines derived from various lymphoid malignancies were determined by Northern blot analysis as described in Materials and Methods. Filters were stained with methylene blue to visualize 18S and 28S ribosomal RNAs (28SrRNA was indicated at the bottom panel). (b) Expression of Chk2 and β-actin proteins in the indicated cell lines was determined by anti-Chk2 or anti-β-actin immunoblotting of whole-cell lysates (equal protein amounts) from the respective cell lines as described in Materials and Methods

Full size image

Decreased expression of Chk2 mRNA in HL cells would presumably result in the decreased expression of Chk2 protein. To verify this possibility, Chk2 protein expression was analyzed by anti-Chk2 immunoblots of whole-cell lysates. As shown in Figure 1b, Chk2 protein expression levels correlate generally with Chk2 mRNA levels in the cell lines examined, and decreased expression of Chk2 protein was observed in all three HL cells compared to other cell lines. It is possible that downregulation of Chk2 expression may be involved in the resistance to apoptosis and the occurrence of aneuploidy frequently observed in HLs (see Discussion).36

Effect of HDAC inhibitors (TsA and SB) on Chk2 expression in HL cells

In order to clarify the mechanisms of altered Chk2 gene expression in HL cells, we tested the effect of the HDAC inhibitors TsA and SB on Chk2 expression in HL cells. HL cell lines (L428, KM-H2 and HDLM2 cells) and several other cell lines (KM3 and RL cells) were treated with either TsA or SB for 2–3 days or left untreated, and the levels of Chk2, p53 and Wip1 mRNA expression were determined by Northern blot analysis. Intriguingly, treatment of HL cells with SB resulted in increased expression of Chk2 mRNA, but to levels lower than those observed in untreated KM3 and RL cells (Figure 1a and 2a, data not shown). Treatment of L428 and KM-H2, but not HDLM2 cells, with TsA also enhanced Chk2 expression to lesser extents compared to that with SB, indicating differential sensitivities of the respective HL cells toward TsA and SB. The results are consistent with the idea that hypoacetylation of histones is responsible for the downregulation of Chk2 expression in HL cells. On the other hand, treatment of HL cells with TsA or SB caused a decrease in p53 and Wip1 mRNA (Figure 2a), indicating that some form of epigenetic regulation is ultimately responsible for p53 and Wip1 expression in HL cells. These results suggest that aberrant gene expression of Chk2, p53 and Wip1 in HL cell lines is due, at least in part, to epigenetic regulation, that is, acetylation of histones.

Figure 2
figure 2

Expression of Chk2 mRNA and protein in HL cells (L428, KM-H2 and HDLM2) cells after treatment with TsA or SB. (a) HL cells (L428, KM-H2 and HDLM2 cells) were treated or not treated with TsA or SB for 2–3 days as described in Materials and Methods. Expression of Chk2, p53 and Wip1 mRNAs was determined by Northern blot analysis as described in Materials and Methods. Filters were stained with methylene blue to visualize 18S and 28S ribosomal RNAs (28SrRNA was indicated at the bottom panel). (b) HL cells (L428, KM-H2 and HDLM2 cells) were treated or not treated with TsA or SB for 2–3 days as described in Materials and Methods. Expression of Chk2 and β-actin proteins in the respective cells under the indicated conditions was determined by anti-Chk2 or anti-β-actin immunoblotting of whole-cell lysates (equal protein amounts) as described in Materials and Methods

Full size image

We next examined expression of the Chk2 proteins in the three HL cell lines following treatment with TsA or SB (Figure 2b). Treatment of L428 and KM-H2 cells with TsA or SB resulted in augmented expression of Chk2, as determined by immunoblot analysis (Figure 2b). Treatment of HDLM2 cells with SB, but not TsA, also caused an increase in Chk2 expression. Thus, in HL cell lines treated with these inhibitors, protein expression correlates with expression of each mRNA. This suggests that expression of these proteins is regulated, at least in part, by epigenetic mechanisms at the transcriptional level.

Effect of DNA methyltransferase inhibitor 5Aza-dC on Chk2 expression in HL cells

We next examined the effect of the DNA methyltransferase inhibitor 5Aza-dC on Chk2 expression (mRNA and protein) in HL cells (Figure 3a). To this end, HL cell lines (L428, KM-H2 and HDLM2 cells) were treated with 5Aza-dC for 2–3 days or left untreated, and the levels of Chk2 mRNA and Chk2 protein were determined by Northern blot and immunoblot analyses, respectively. As shown in Figure 3a (upper panel), treatment of HL cells with 5Aza-dC resulted in increased expression of Chk2 mRNA, but to levels lower than those observed in untreated KM3 and RL cells (data not shown), suggesting that DNA methylation is also involved in the downregulation of Chk2 expression in HL cells. In agreement with Chk2 mRNA expression, treatment of HL cells with 5Aza-dC also resulted in augmented expression of Chk2 protein, suggesting that expression of Chk2 protein is regulated, at least in part, by epigenetic mechanisms at the transcription level.

Figure 3
figure 3

Methylation status of the Chk2 gene in HL cells. (a) HL cells (L428, KM-H2 and HDLM2 cells) were treated or not treated with 5Aza-dC for 2–3 days as described in Materials and Methods. Expression of Chk2 mRNA was determined by Northern blot analysis, and expression of Chk2 and β-actin proteins was determined by anti-Chk2 or anti-β-actin immunoblotting of whole-cell lysates (equal protein amounts) as described in Materials and Methods. (b) Overall methylation pattern of the Chk2 gene was examined by Southern blot analysis using full-length cDNA probe for Chk2 as described in Materials and Methods. The molecular marker is shown to the right. (c) Methylation status within the CpG-rich region of Chk2 gene was determined by MSP as described in Materials and Methods. Locations of the CpG-rich region and the MSP primer sets (primer sets 1 and 2) are indicated (upper panel). Bisulfite-treated DNAs from L428 and KM3 cells were amplified using two unmethylated (U) and methylated (M) DNA-specific primer sets. Amplification with the primer set 2 (U and M) is indicated (lower panel)

Full size image

Methylation status of Chk2 gene in HL and other malignant lymphoid cell lines

The effects of 5Aza-dC on Chk2 gene expression in HL cell lines suggest that methylation of cytosine residues at CpG dinucleotides in the Chk2 gene may be involved in the regulation of Chk2 gene expression in these cells. To test this possibility, we examined overall methylation patterns in the Chk2 gene in HL (L428 and KM-H2 cells) and other cell lines (KM3 and RL cells). Genomic DNA from each cell line was digested with an isoschizomeric restriction enzyme pair; HpaII, which is methylcytosine-sensitive, and MspI, which is methylcytosine-resistant, or with HindIII as a control, and Southern blot analysis was performed as described in Materials and Methods. As expected, the genomic DNA from all the four cell lines digested with MspI or HindIII exhibited identical band patterns, while that digested with HpaII exhibited a different pattern depending on the cell line (Figure 3b). It was also found that the overall methylation status of the Chk2 gene in HL and other cell lines does not correlate with the expression levels of Chk2 mRNA.

Database analysis revealed the presence of a CpG-rich region upstream of a putative transcription initiation site of Chk2 gene (Figure 3c). Therefore, we examined the methylation status within this CpG-rich region in L428 and KM3 cells using a methylation-specific PCR (MSP) method employing two MSP primer sets (primer sets 1 and 2) (Figure 3c). Genomic DNA from L428 and KM3 cells produced a strong PCR product with unmethylated sequence-specific primers, but not with the methylated sequence-specific primers (primer set 1). MSP analysis of the genomic DNA from both L428 and KM3 cells using primer sets 2 also amplified only the unmethylated sequence (data not shown). Thus, the CpG-rich region of Chk2 in both L428 and KM3 cells appears to be unmethylated.

Effect of 5Aza-dC TsA and SB on histone acetylation and methylation at the Chk2 promoter in L428 and KM3 cells

To examine the acetylation status of the histones H3 and H4 associated with Chk2 gene, in particular the Chk2 promoter, in HL cells, we performed Chip using polyclonal antibodies against acetylated histone H3 and acetylated histone H4. We examined untreated KM3 cells, untreated L428 cells, and TsA- or SB-treated L428 cells as described in Materials and Methods. As shown in Figure 4, we observed that histone H3 (and histone H4) associated with the Chk2 promoter was hypoacetylated in untreated L428 cells compared to untreated KM3 cells. Treatment of L428 cells with either TsA or SB resulted in hyperacetylation of both histones H3 and H4, suggesting that decreased expression of the Chk2 gene in the HL cell line L428 is due, at least in part, to hypoacetylation of histones H3 and H4.

Figure 4
figure 4

Levels of acetylated histone H3 and H4 and dimethylated histone H3 lysine 9 at the Chk2 promoter in L428 and KM3 cells after treatment with 5Aza-dC, TsA or SB. L428 and KM3 cells were treated with 5Az-dC, TsA, or SB or left untreated for 2–3 days as described in Materials and Methods. Chip was performed with antidimethyl histone H3 (K9), antiacetyl histone H3, antiacetyl histone H4 or control antibodies, and the immunoprecipitated DNA was subjected to PCR with primers specific to the promoter region of the Chk2 gene as described in Materials and Methods. Chip PCR products were separated in 2% agarose gels, and stained with ethidium bromide. Histograms show quantitation of band intensity of the respective Chip PCR products (lower panel). Data are expressed as the mean±S.D. (relative to the levels in untreated cells (−, mock treated)) in four independent experiments

Full size image

Since it has been shown that methylation of histone H3-lysine (K) 9 is associated with inactive genes, we also examined the methylation status of this histone by Chip assay. Methylation of Chk2 promoter-associated H3-K9 was clearly higher in untreated L428 cells compared to untreated KM3 cells (Figure 4). This suggests that hypermethylation of histone H3-K9 may also be involved in repression of Chk2 gene expression in L428 cells. Accumulating evidence demonstrates an interplay between DNA methylation and histone modification (methylation and acetylation).42 The methyl-CpG binding protein, MeCP2, associates with HDAC complexes, suggesting a mechanism by which histone modification can be induced by changes in DNA methylation.43, 44 Interestingly, treatment of L428 cells with 5Aza-dC resulted in the drastic inhibition of histone H3-K9 methylation and enhancement of histones H3 and H4 acetylation, showing that there is indeed an interplay between DNA methylation and histone modification. Collectively, these results indicate the crucial role of epigenetic mechanisms in downregulating Chk2 gene expression in the HL cell line, L428 cells.

Apoptotic changes induced by γ-irradiation in L428 cells by treatment with SB

Recent studies have shown that Chk2−/− mice and cells have decreased susceptibility to ionizing radiation (IR).45, 46 It has also been reported that HL cells frequently exhibit resistance to apoptosis.36 Therefore, it is possible that the observed alteration in the expression of Chk2 and related genes in HL cells may be involved in apoptosis resistance. To examine this possibility, we examined the effect of SB on the susceptibility of HL cells (L428 and KM-H2 cells) to apoptosis induced by γ-irradiation, as described in Materials and Methods. L428 cells were treated with SB, or left untreated, subjected to γ-irradiation (15 Gy), and labeled 24 h later with PI and annexin V. PI/annexin V dual staining indicated that the percentages of annexin Vhigh PIhigh cells relative to total cells in untreated L428 cells increased from 4 to 14% in response to γ-irradiation, while the percentages in SB-treated L428 cells increased from 5 to 27% (Figure 5). Similar results were obtained when KM-H2 cells were treated with SB (data not shown). These results imply that upregulation of Chk2 and other related genes in HL cells by SB treatment contributes to the increased susceptibility of HL cells to γ-irradiation.

Figure 5
figure 5

Detection of apoptotic changes induced by γ-irradiation in L428 cells treated with SB. L428 cells were treated with SB (final conc. 1 mM) or left untreated for 2–3 days, subjected to γ-irradiation (15 Gy), and subsequently cultured for 24 h. Cells were labeled with FITC-conjugated annexin V mAB and PI, and analyzed by flowcytometry as described in Materials and Methods. Numbers indicate percentages of annexin Vhigh PI high, annexin V high PI low and annexin Vlow PIhigh cells relative to total cells. Essentially identical results were obtained in two independent experiments

Full size image

Discussion

Chk2 kinase plays a central role in signaling in response to DNA damage, functioning downstream of ATM and upstream of p53.2, 3 Following DNA damage, Chk2 is phosphorylated and activated by ATM-dependent and -independent mechanisms,4, 5, 6, 7, 8 and this activated Chk2 then phosphorylates and stabilizes p53.1, 2, 3, 10, 11 This, in turn, leads to induction of the p21 and Wip1 genes, which encode a cyclin-dependent protein kinase inhibitor (CKI) p21, and the oncogenic nuclear protein phosphatase Wip1, respectively.47, 48, 49 Interestingly, Wip1 has been shown to mediate negative feedback regulation of p38–p53 signaling in response to DNA damage.50 In addition, it has been well documented that most if not all molecules involved in DNA damage response signaling are tumor suppressors, and mutations of these genes have been implicated in a wide variety of tumors. In fact, germ-line mutations of the Chk2 gene have been identified in a subset of families affected by LFS,15, 16, 17 although somatic Chk2 mutations are rather rare, particularly in lymphoid malignancies.23, 24, 25

In our present study, we found that cell lines from HLs exhibit drastically decreased expression of the Chk2, but not p53 and Wip1, compared to cell lines from other lymphoid malignancies. We found that this decreased expression was due, at least in part, to altered epigenetic regulation (Figures 1 and 2). In particular, downregulation of Chk2 gene expression was shown to involve hypoacetylation of histones H3 and H4, hypermethylation of histone H3-lysine (K) 9 and DNA hypomethylation at an as-yet-unidentified site(s) outside the CpG-rich region in the Chk2 promoter (Figures 3 and 4). Although a recent study has shown that aberrant post-transcriptional regulation of Chk2 expression may occur in a subset of aggressive non-HLs,32 we believe our data represent the first example of altered epigenetic regulation of Chk2 gene expression in cells from malignant tumors.

HLs are characterized by the presence of malignant mononuclear H/RS cells, surrounded by reactive cells attracted by cytokines and chemokines, which are abundantly produced by H/RS cells.51 It has been reported that constitutive activation of nuclear factor (NF)-κB and AP-1 are also implicated in HLs,52, 53, 54 and that genetic and/or epigenetic alterations (e.g. mutation, splicing variant and DNA methylation) in genes encoding cell-cycle regulators, including p16INK4a, p14ARF and IκBα, exist in H/RS cells from HLs.33, 34, 35, 36 Furthermore, a recent study using tissue-microarrays has revealed that Cyclin E, Cdk2, 6, Stat3, Hdm2, Bcl2, Bcl-XL, Survivin, molecules involved in cell-cycle regulation and prevention of apoptosis, are overexpressed in HLs, explaining their high proliferative activity and resistance to apoptosis.36 Intriguingly, polymorphic mutations of the ATM gene have recently been identified in childhood Hodgkin's disease.55 These mutated ATM genes encode functionally abnormal ATM proteins, suggesting the involvement of polymorphic variations of the ATM gene in the pathogenesis of childhood HD, presumably due to a defect in cell-cycle checkpoint regulation. Our findings, that altered epigenetic regulation of Chk2 gene in HL cells results in the downregulation of Chk2 kinase and abrogates DNA damage response signaling in the cells (Figure 5, see below), suggest that this novel molecular feature of HLs that may also be useful for the diagnosis of the disease. In fact, downregulated expression of Chk2 and abrogation of DNA damage response could also account for the resistance to apoptosis and aneuploidy occasionally observed in HLs.

Our results demonstrate that downregulation of Chk2 gene expression in HL cell lines is due partly, if not entirely, to altered epigenetic regulation of the Chk2 gene, that is, hypoacetylation of histones H3 and H4, hypermethylation of H3-K9 and hypomethylation of DNA (Figures 2, 3a and 4). Although overall DNA methylation analysis of the Chk2 gene and MSP analysis of the CpG-rich region of Chk2 failed to detect a DNA methylation site(s) within the Chk2 gene that differs between HL and cells from other lymphoid malignancies (Figures 3b and c), our results with 5Aza-dC (see Figures 3a and 4) nevertheless indicate the involvement of DNA methylation in decreased Chk2 expression in HL cells. Further study will be required to identify potential crucial cytosine residues within the Chk2 gene. Our results with 5Aza-dC also provide an example of the interplay between DNA methylation and histone modification (methylation and acetylation)42 (Figure 4). This interplay can be assumed to be mediated by a molecular mechanism whereby the methyl-CpG binding protein, MeCP2, associates with HDAC complexes.43, 44 However, a recent study using a genetic approach has indicated that histone modification can occur independently of DNA methylation, and that DNA methylation serves to ‘lock in’, rather than initiate gene repression.56 Further study will be required to clarify this issue. Moreover, it has been appreciated that epigenetic regulation of gene expression can be affected by environment surrounding cells. Thus, the epigenetic regulation in particular cells in vitro and in vivo may be different. Since our present study is restricted to established HL cell lines, it will be of importance to examine the epigenetic regulation of the Chk2 gene by utilizing fresh specimens from patients with HLs.

It has been shown that targeted disruption of the Chk2 gene results in decreased susceptibility of cells in response to IR, and that HL cells occasionally exhibit resistance to apoptosis.36, 45, 46 Consistent with these observations, we found that upregulation of Chk2 in HL cells (L428 and KM-H2 cells) following treatment by the HDAC inhibitor SB resulted in increased susceptibility of the cells to IR (Figure 5). At present we do not know to what extent SB-induced Chk2 expression contributes to this increased susceptibility. Future studies examining the effects of ectopically expressed Chk2 on the susceptibility to IR in HL cells might address this question. Importantly, our results may also provide some insight into the potential therapeutic use of HDAC inhibitors against malignant tumors, especially on the combined treatments (e.g. HDAC inhibitor and IR) for patients with HLs.

Materials and Methods

Cell lines and culture conditions

Three HL cell lines and six other lymphoid malignancies cell lines were analyzed. The HL cell lines were L428, HDLM2 (both nodular sclerosis),57 and KM-H2 (mixed cellularity),57 and the other cell lines were KM3 (pre-B lymphocytic leukemia),58 RL (non-HL),59 FL318 and FL518 (follicular lymphomas),60 Black 93 (Burkitt's lymphoma),61 and Jurkat (T cell leukemia).62 All cell lines were maintained in RPMI1640 supplemented with 10% (v/v) fetal calf serum (FCS). L428 cells were treated with 5Aza-dC (5Aza-dC, Sigma) (final conc. 5 μ M), TsA (Sigma) (final conc. 300 nM), or SB (Aldrich) (final conc. 1 mM) for 2–3 days.

Northern blot analysis

Cells were harvested and total RNA was prepared using the ISOGEN RNA preparation kit (Wako). Total RNA (10 μg) was electrophoresed in 1% agarose formaldehyde gels and transferred onto nylon membranes (Nytran N, Schleicher & Schuell) as described previously.63 Probes were labeled with [α-32P]dCTP (Amersham, 3000 Ci/mmol) using the Multiprimer labeling kit (Amersham) and were hybridized using PerfectHyb (TOYOBO) following the manufacturer's instructions. Specific activity was approximately 2 × 106 cpm/ng for each of the probe DNAs. 28S rRNA was visualized by staining filters with methylene blue. The probe DNAs were prepared as follows: Chk2,1 1.7 kbp EcoRI–XhoI fragment from pcDNA3-Chk2; p53,64 1.8 kbp BamHI fragment from pCMV-p53; Wip1,50 1.8 kbp BglII–SphI fragment from pCMV-Wip1.

Southern blot analysis

Genomic DNAs were prepared from the four cell lines (L428, KM-H2, RL and KM3) by phenol–chloroform extraction. To detect methylation sites within genomic DNA, genomic DNA (10 μg) from each cell line was digested with HpaII (New England Biolabs) or MspI (TAKARA). Both restriction enzymes recognize the same sequence ‘CCGG’, but HpaII cannot digest the methylated sequence, while MspI can.65, 66 HpaII- or MspI-digested DNAs were electrophoresed in 0.8% agarose gels, and transferred onto nitrocellulose membranes (PROTRAN BA 85 CELLULOSENITRAT, Schleicher & Schuell). Probes were labeled as described in Northern blot analysis. After hybridization, the membranes were washed with 5 × SSC/0.1% SDS at 65°C.

Methylation-specific PCR analysis

To perform MSP,67 genomic DNA (2 μg) was treated with sodium bisulfite using the DNA Modification Kit (CpG Genome, Intergen), following the manufacturer's instructions. Bisulfite converts unmethylated cytosines, but not methylated cytosines, to uracil. The sequences of the MSP primers were as follows. Primer set 1: unmethylated DNA (U); 5′-GTTTTTTTTTTTTGTAGGTTAGATTTTGAT-3′ and 5′-CTTCAACCTTATAAACTAATACAAACAACA-3′. methylated DNA (M); 5′-GTTTTTTTTTTTCGTAGGTTAGATTTCGAC-3′ and 5′-CTTCAACCTTATAAACTAATACGAACGACG-3′. Primer set 2: (U); 5′-TTGTTTGTGATGTAGTATTGTAGTTTAGTG-3′ and 5′-ACAAATAACCACAACTAAATAAACCACCAA-3′. (M); 5′-TCGTTTGCGACGTAGTATCGTAGTTAGCG-3′ and 5′-ACGAATAACCACGACTAAATAAACCGCCGA-3′. The sequences recognized by these primers were located within a likely methylation site, a CpG-rich region upstream of the putative transcription initiation site of the Chk2 gene. Following PCR, the respective samples were fractionated on 2% agarose gels and visualized under UV.

Protein preparation and immunoblotting

Cells were harvested and solubilized with lysis buffer (50 mm Tris-HCl (pH 7.4), 0.5% (v/v) NP-40, 150 mm NaCl, 5 mm EDTA, 50 mm NaF, 1 mm Na3VO4, 1 mm phenylmethyl sufhonyl fluoride (PMSF), 10 μg/ml leupeptin and 10 μg/ml aprotinin), and cell lysates were prepared by centrifugation at 12 000 × g for 15 min at 4°C. Amounts of proteins in the respective cell lysates were quantitated by the Bio-Rad DC protein assay kit (Bio-Rad). Cell lysates containing equal amounts of proteins were separated by SDS-PAGE (10% PAG), and transferred onto PVDF membrane filters (Immobilon, Millipore). The membranes were immunoblotted with rabbit polyclonal anti-human Chk2 (HF, NK, TK, YM, unpublished data), anti-p53 (DO1, Santa Cruz), or mouse monoclonal anti-β-actin (AC-15, Sigma-Aldrich) antibodies, and bound antibodies were visualized with HRP-conjugated goat anti-rabbit IgG or goat anti-mouse IgG antibodies (Bio-Rad) using a chemiluminescence reagent (Western Lightning, Perkin Elmer Life Sciences) following the manufacturer's instructions. Rabbit polyclonal anti-Chk2 antibody was raised against peptides corresponding to amino acids 523–543 of human Chk2.

Chip assay

In total, 1 × 106 cells were treated with 1% formaldehyde for 15 min at 37°C to crosslink proteins to DNA. Subsequently, chromatin was solubilized and subjected to sonication to obtain DNA fragments with an average size of 750–1000 bp. Chip assays were performed as described previously68 with an equal amount (5 μl) of antiacetyl histone H3, antiacetyl histone H4, antidimethyl histone H3 (K9) antibodies (Upstate Biotechnology) or normal rabbit IgG (Capel) as a control. The immunoprecipitated DNA was analyzed by PCR with primers specific to the Chk2 promoter region. The sequences of the primers utilized were 5′-GACGGAGTTTCACTATGTTGGCC-3′ and 5′-CTGCCATGAGACTGCTGAGCCTCAACAT-3′, respectively.

Analysis of apoptosis

Cells treated with or without SB were subjected to γ-irradiation (15 Gy) and cultured for 24 h. Cells were harvested and labeled with annexin V and PI using the TACS Annexin V-Apoptosis Detection kit (R&D systems) following the manufacturer's instructions and subsequently analyzed by a flow cytometry (FACSCalibur, Becton Dickinson). The percentages of cells in each fraction were analyzed using the software CellQuest.