Original Paper | Published:

Methylation of the asparagine synthetase promoter in human leukemic cell lines is associated with a specific methyl binding protein

Abstract

We have examined the methylation profiles of the asparagine synthetase (ASY) promoter in a number of human leukemic cell lines in relation to their asparagine (ASN) requirements in vitro. Cells in which the promoter is highly methylated are auxotrophs and express ASY at very low levels. Electromobility shift assays (EMSA) of nuclear extracts with oligomers from the promoting region show, in addition to recognized transcription factor binding, a novel methyl binding protein specific for a 12 base consensus sequence, which includes a single methylated CpG. This sequence overlaps that of the amino-acid response unit of the ASY promoter, which is activated byATF4 and C/EBP. Competition by the methyl binding protein could account for the observed failure of the methylated promoter to bind these transcription factors and consequently, although other mechanisms can also be operative, for the specific repression of the gene. The ASY methyl binding protein (ASMB) is present in leukemic lymphoid and myeloid cells irrespective of their methylation status, and in normal lymphocytes after phytohemagglutinin stimulation. It has been purified by affinity chromatography and has a molecular size of 40 kDa in 10% SDS-polyacrylamide gels.

Introduction

Asparagine synthetase (ASY) is a housekeeping enzyme expressed constitutively in mammalian cells. However, a number of types of leukemic cells in rodents and humans are auxotrophs, in which the parent gene (ASY) is not expressed, forming the basis for the use of L-asparaginase in therapy (Broome, 1968). The gene has been cloned from animal and human sources (Andrulis et al., 1987; Greco et al., 1989) and both cis and trans transcriptional control has been identified. Guerrini et al. (1993) demonstrated a region from −70 to 62 that, by reporter assay, responded to amino-acid deprivation. Zhong et al. (2003), using Hep2 cells, have shown induction of ASY by deprivation not only of asparagine (ASN) but also of multiple individual essential amino acids, constituting an amino-acid response pathway. This pathway is also activated by glucose deprivation and the endoplasmic reticulum stress response. All were shown to operate through the same response unit, which has been identified in a highly conserved region of the promoter. It consists of two response elements, NSRE-1 extending from −68 to −60, an 11 base spacer that can be varied in sequence but not in number, and a further response element NSRE-2, extending from −48 to −43. Both elements are necessary for function. Siu et al. (2002, 2001) identified transcription factors that bind to NSRE-1, namely ATF 4 and members of the C/EBP group, particularly the β isoform, both occurring in the same complex in mobility shift assays. Transcription factors acting on NSRE-2 have not yet been identified. Taken together, these findings suggest the binding of transcription factors in separate complexes at each response element, linked by one or more molecules extending across the obligatory 11 base spacer.

These studies converge with our own on the nature of ASN dependence in lymphoma cells. Cells of the mouse lymphomas 6C3HED and 1543, which are ASN dependent, are highly methylated in the ASY promoting region and do not express ASY, in contrast to variants that are ASN independent and actively express the gene (Peng et al., 2001). The present paper extends these findings to a series of human leukemic cell lines in vitro. Two lines, 1873 and 1929, are auxotrophs with the ASY promoter highly methylated, while prototrophic variants are at least partially unmethylated, as are other lines that are intrinsically ASN independent. On examining the pattern of nuclear factor binding to oligomers of the promoting region, distinct differences are found dependent on their methylation. Associated with this region, a methyl binding protein is demonstrable, which differs from those studied by Bird and co-workers (Bird and Wolffe, 1999), and others (Jost and Hofsteenge, 1992; Ehrlich and Ehrlich, 1993), notably in terms of the high degree of sequence specificity it shows for its DNA binding site. This binding site overlaps the ASY response unit, particularly the first response element and correlates with the inhibition of binding of ATF4 and C/EBP transcription factors when DNA is methylated.

Results

The various leukemic cell lines which have been studied proliferate in Dulbecco's modified Eagle's Medium (DMEM) containing 10% fetal calf serum, supplemented with L-ASN (50 mg/li) in the case of 1873 and 1929. With these two lines, the subject of most experiments to be described, proliferation ceases abruptly in the absence of added ASN, but apoptosis occurs much more slowly (Figure 1a and b). With 1929 cells, viability drops off sharply (to 60%) at 6 days, but 1873 cells survive considerably longer and even at 26 days 10% remain viable. This contrasts with the rapid loss of viability of ASN-dependent mouse lymphoma cells under similar conditions (Broome, 1963).

Figure 1
figure1

Proliferation assays of leukemia cells in Dulbecco's medium (D) and with added L-asparagine (50 mg/li) (D+ASN) measured by the tetrazolium blue reaction in 96-well plates. Points shown are the average of three determinations. (a and b) Viability of cells in unmodified DMEM is shown by trypan blue exclusion. Cell-free controls of D and D+ASN are included in A. (c) Proliferation of an ASN-independent variant (1873 IND), and the intrinsically independent line REH, in an ASN-depleted medium is shown. (d) Comparison of ASY mRNA levels in leukemic cells of various types grown in D+ASN, measured by real-time RT–PCR. The ASN-dependent cells (1873 and 1929) had a markedly lower level than all others (prototrophs). Other experiments showed that for protrophs, mRNA levels did not change with ASN depletion of the medium

The expression of ASY correlated with the ability of cells to grow in the absence of ASN, with the lowest level of expression occurring in the auxotroph 1929 (Figure 1d). Various lines, intrinsically ASN independent had a high level of expression. An ASN-independent variant of 1873 (IND) (Figure 1c) had about twice the level of the original wild type. Consistent with findings on the methylated ASY promoter in mouse lymphomas, 5-aza-2′ deoxycytidine produced marked increases in expression in both 1873 and 1929 cells. Maximal results were seen at 48 h with a concentration of 10 μ M (Figure 2), but after this time the agent proved toxic.

Figure 2
figure2

Effect of 5-aza-2′ deoxycytidine on expression of ASY in 1929 and 1873 wild-type cells, measured by real-time RT–PCR. Results shown are averages of four separate cultures

These results also correlated with methylation profiles of the 5′ region of the gene, specifically in bases −249 to +25 that contained 24 CpG groups (Figure 3a). DNA clones showed essentially complete methylation in 1873 and 1929 cells, in contrast to an ASN-independent variant (1873 IND), where only two of nine clones were fully methylated. REH cells, which from standard cultures, showed five of 10 DNA clones totally methylated, proved by cellular cloning (expansions from 24 cells) to contain individually both a methylated and an unmethylated gene. In Raji, NB4 and Jurkat cells that were intrinsically ASN independent, no methylation was observed. The region of the ASY gene examined included promoter elements studies by Guerrini et al. (1993) (−151 to −31), which contained a number of potential sites for transcription factor binding (Biosite, Wolfenbuttel, Germany) Three sets of probes were prepared from the region and nuclear extracts were examined by electromobility shift assays (EMSA). The distal probe (−151 to −104) possessed only poor and indistinct binding activity, the middle probe (−108 to −71) that contained sites for SP1 and WT1, bound SP1, but not WT1, as shown by competition and supershift assays. The EMSA pattern did not change with methylation of the oligomer, consistent with other reports of binding by SP1 (Holler et al., 1988; Ohtani-Fujita et al., 1993).

Figure 3
figure3

Methylation profiles of ASY promoting regions in leukemic cells. (a) Diagram and numeration of potentially methylated CpG sites in the region examined. (b) Identical methylation profiles are seen in wild-type 1873 and 1929 cells. (c) Methylation profile in the variant 1873 IND showing extensive but incomplete demethylation. □ – mutation G → A at site #19. This mutation coupled with isolated methylation at sites #5 and 6 forms a frequent pattern in clinical samples of ALL

The proximal probe (−75 to −33) proved to be the most interesting. Using an unmethylated oligomer of 42 bases, EMSA showed three principal bands: (1) close to the origin; (2) mid region, relatively sharp; and (3) distal, heavier and more diffuse (Figure 4). By contrast, using a methylated probe, the mid-region band was markedly diminished or absent, and an additional band was found, with greater mobility then the others; this methyl binding protein was designated ASMB. Neither the methylated nor the unmethylated oligomers competed with each other. Binding occurred only when both strands were symmetrically methylated.

Figure 4
figure4

EMSA pattern of nuclear extract of 1873 using proximal ASY promoting region (−75 to −33) oligomer (Table 1). The unmethylated oligomer (R10) shows three major bands; the first two are much diminished with methylation of CpG at position #16 (R15) and a methyl binding protein, ASMB, becomes apparent. Substantial modification of the native oligomer sequence at 5′ and 3′ ends (R15A) retains ASMB binding capacity but other binding characters are changed. The shorter methylated oligomer of 22 bases (R15B) binds protein, but with different mobility from the 42 base oligomer. Other methyl protein binding oligomers (underlined sequences in Table 1b), MM2 and MKL (Kaiso) do not bind ASMB, even when, as shown, they form part of a 42 base sequence

Further studies by EMSA, confirmed by competition experiments, showed that the ASMB was present in the T cell leukemia Jurkat, and in normal ficol-hypaque separated peripheral blood lymphocytes when stimulated by phytohemagglutinin (48 h), but it was not found in unstimulated cells. A weak expression was found in the myeloid leukemia K562 and in Hep2. No expression was found in HeLa or two other carcinoma cell lines (MCF7 and SKBR3).

Although three CpG sites were present in the oligomer sequence, binding activity was related to only one of these (#16 in Figure 2a and oligomer R15, Table 1). In addition, a core sequence of 12 adjacent bases was also necessary for binding, but outside this, replacement of runs of bases at 5′or 3′ ends of the oligomer, while altering the general binding pattern (for example oligo R15A), did not prevent the binding of ASMB (Table 1 and Figure 4). On the other hand, the oligomer length mattered. A methylated oligomer of 22 bases with the native ASY sequence (R15B) had methyl binding activity and showed competition with the longer oligomer, but the mobility of the bound protein was distinctly different.

Table 1 (a) Sequence and variations in proximal promoting region oligomer (−75 to −33) of ASY with their effect on binding ASMBa, and (b) sequences of various methylated oligomers with protein binding activity are shown incorporated into the 42 base sequence of the ASY oligomerb

Numerous substitutions of bases in the central region of the 42 base oligomer (Figure 5 and Table 1) defined the specificity to a consensus sequence of AANTTCYMGCRC (Table 1, showing a numeration of bases). Substitution of C by A or T in the position 6, or A by G in position 12 invariably resulted in a much heavier band. However, when A in the fifth position was replaced by G, a heavy band with altered motility resulted (Figure 5), and so too when C in the ninth position was replaced by A.

Figure 5
figure5

EMSA pattern of nuclear extract of 1929 to show the effect of single-base substitutions of R15. These produce diminution or absence of the characteristic methyl binding protein, ASMB (positions 4, 5, 7 and 8) or bind protein or with different mobility (position 5). A correlation is found between with loss of band 2 and diminished intensity of band 1, as defined in Figure 4

These observations indicate that the ASY sequence is less than optimal for binding the protein, and also that other binding proteins (or complexes) with different precise specificities exist in the nuclear extracts. It was observed that in all cases, when the characteristic methyl binding protein, ASMB, was observed in the gels, binding of other proteins in the band near the origin (1) and particularly at the mid part of the gel (2) was absent or greatly reduced.

Siu et al. (2002, 2001) have shown that in Hep 2 cells, the response element NSRE-1 binds ATF4 and members of the C/EPB families, both of which are necessary, but not sufficient for the activation of ASY (Zhong et al., 2003). The general patterns of electromobility shifts of nuclear proteins in our preparations with oligomers of native sequence differed from these observations, but we used longer oligomers (42 bases as opposed to 26). Leukemic cell extracts gave a supershift with ATF4 antibody in band 2 (Figure 6). In the region closest to the origin (band 1), supershifts were found with antibodies to α, β and γ isoforms of C/EPB, and to the greatest degree with δ, but not with ɛ. This differed from the predominant β form in hepatic cells, but the various isoforms are known to be tissue related (Ramji and Foka, 2002). Methylation of the oligomer, as has been described, abolished (or severely reduced) binding to bands containing these factors.

Figure 6
figure6

EMSA pattern of unmethylated oligomer R1O, identifying ATF4 and C/EBP transcription factor binding by supershift. A nuclear extract of 1929 was used. A supershift of band 2 was produced with antibody to ATF4. A supershift with antibody to C/EBPδ is found in band 1 and less marked supershifts with α, β and γ. No supershift is produced with C/EBPɛ

Experiments performed by chromatin immunoprecipitation (CHIP) examined the binding of ATF4 to the ASY promoter. Preparations obtained from wild-type cells of 1929 and 1873, in both of which the promoting region was totally methylated, gave no detectable signal for binding. By contrast, a subline of 1929 (demethylated and growing in co-culture) gave a signal that compared with the corresponding input DNA in a ratio of 1.4 : 1. Similarly, products of demethylated cells of two sublines derived from 1873 (1873 IND and co-cultured 1873 SN3) gave ratios of 0.91 : 1 and 0.92 : 1, respectively. Raji gave a ratio of 0.83 : 1.

Since a number of proteins have been shown to bind methylated DNA, it was necessary to show that their binding characteristics differed from those of ASMB by direct comparison. Bird and co-workers (reviewed by Bird and Wolffe, 1999) have demonstrated several categories of methyl binding proteins. Of particular present interest, it was shown that when a number of oligomers containing a single methylated CpG were examined, specifically the 27 base oligomer MM2, they were found to bind to a protein, MeCP2, and binding occurred without nucleotide sequence specificity. A further group of proteins, MBD 1, 2 and 4 were later identified and found to bind similarly to singly methylated oligomers (Hendrich and Bird, 1998).

A comparison has been made of the binding properties of ASMB to these other proteins. Direct experiments show that neither MM2 (nor the unmethylated MM1) produce any band similar to ASMB on EMSA, even when the 27 base MM2 oligomer is extended to 42 by the addition of sequences from the terminal parts of our probe (Figure 4). Furthermore, 100-fold excess of MM2 failed to compete with ASMB for the latter's specific binding site (oligomer R15). Although Western blots show the presence of MeCP2 and MBD 1–4 in nuclear extracts of the leukemic cells, assays with antibodies to these proteins showed no supershift of the ASMB band.

A further protein, Kaiso (Daniel et al., 2002) has been described, which recognizes a consensus sequence, ‘metastatin’ that is methyl CpG dependent (shown in Table 1). Oligonucleotides containing this methylated site, either in the 22 original base sequence or when incorporated into a 42 base oligomer with extension at either end from our probe sequences, did not produce any band on EMSA comparable to ASMB (Figure 4). Similar negative findings were obtained with oligonucleotides containing the sequence MBDP-1, which has been shown to have methyl-specific protein binding activity (Khan et al., 1988; Ehrlich and Ehrlich, 1993).

To further characterize ASMB, we undertook its preliminary purification by affinity chromatography using the methylated oligomer R15, covalently linked to sepharose. Firm binding of the protein occurred, which was not displaced by washing buffer at PH 7.5 (containing poly dA/dT) or by excess of unmethylated oligomer (R10), but elution readily occurred with free methylated oligomer (R15). Electrophoresis of the eluate on 10% PAGE–SDS gels yielded a single band of approximately 40 kb size.

Discussion

The concept that methyl binding proteins act by preventing the action of transcription factors has had several iterations. The work of Jost (reviewed by Jost and Saluz, 1993) shows that in immature chicks the yolk protein vitellogenin is not expressed in the liver; its promoting region is methylated and binds to a modified and phosphorylated form of histone HI. On administration of estradiol, demethylation occurs, leading to gene expression. Ehrlich and co-workers have shown that a protein designated MBDP-1 obtained from the placenta, binds to a number of related viral and animal DNA sequences, some in promoter regions (Khan et al., 1988, Ehrlich and Ehrlich, 1993). Although the protein shows a degree of specificity for 5-methyl cytosine, this can in most circumstances be replaced by thymidine. MBDP-1 has been identified with RFX, a dimer of 180 kDa that regulates the expression of HLA II class genes, through binding to the so-called ‘x box’, a 14 bp consensus sequence (Zhong et al., 1993). This sequence is present in numerous sites, notably in collagen, apolipoprotein and in viral promoters. In a rat fibroblast system that does not express collagen, the α 2(1) promoter is methylated, and it binds RFX (Sengupta et al., 1999). The effect of binding, particularly its relation to specific transcription factors, has not been determined. Recently, the zinc-finger protein, Kaiso, which is found as a partner to C120 catenin, has been shown to have dimorphic DNA binding sites. One of these binds in a consensus sequence requiring two (optimally three) methylated CpG groups (Daniel et al., 2002).

The properties of our protein do not appear related to these. Clearly, the MBDP-1 of Ehrlich and Ehrlich (1993) differs as thymidine cannot replace methylated CpG in our protein and its size (40 kDa) is clearly very much smaller than the 140 kDa of the RFX dimer (Zhong et al., 1993). In a similar way, the histone of Jost and Hofsteenge (1992) differs on the grounds of lack of sequence specificity and size (21 kDa). Kaiso differs (inter alia) in terms of requiring two methylated CpGs (Daniel et al., 2002).

As distinct from these studies, particular interest in methyl binding proteins and their effects has been derived from the work of Bird and colleagues. Initially a protein MeCP1, of 200–400 kDa, was identified (Meehan et al., 1989), which combined with methylated oligomers (containing 12 or more symmetrically methylated CpGs) but did not otherwise have sequence specificity. This has subsequently been shown to be a complex, which contained MBD2 and the nucleosome remodeling and histone deacetylase complex NuRD. Gene repression was associated with histone deacetylation and with methylation of lysine 9 of histone H3 (Feng and Zhang, 2001). Subsequently the protein, Me CP2 (Nan et al., 1993), which binds oligomers containing a single methylated CpG, was found to be associated physically with histone deacetylase (s), (Jones et al., 1998) with the co-repressor Sin 3A and by tethering a histone methyl transferase, to methylate lysine 9 of H3 (Fuks et al., 2003). There is thus increasing evidence that methyl binding proteins employ related pathways through multiple linked histone modifications to produce chromatin remodeling and gene repression (Wade, 2001; Jones, 2002; Bird, 2002). Four proteins, MBD 1–4 (Hendrich and Bird, 1998), are related to MeCP2 in having similar methyl binding domains, and require only a single methylated CpG. It has recently been shown that on an individual basis, these proteins can bind preferentially to particular methylated promoters; however, with other genes they bind jointly as part of a large repressive complex (Balestar et al., 2003). Although direct comparison shows that oligomers binding to this series of methyl binding proteins do not bind ASMB, Western Blots show that they are present in leukemic cells, and the ASY promoting region contains up to 24 CpG groups susceptible to methylation. EMSA studies, however, have not shown their binding to the oligomers we have studied.

The distinctiveness of ASMB is that it has not only strict requirement for a binding site with a single methylated CpG but also in the adjacent sequence specificity as shown below (Figure 7). This binding sequence overlaps the first response element (NSRE-1) of the ASY gene and extends close to the beginning of the second (NSRE-2). Although additional inhibitory actions of ASMB are by no means excluded, it is postulated that competition for this site with ATF4 and C/EPB is important in gene inhibition. In the case of some base variants of our oligomer, in which a band of binding protein was not observed on EMSA there was often partial through incomplete displacement of ATF4. This could be due to competition from relatively weak binding, which was insufficiently strong to lead to a visible yield of ASMB on the gels. The 11 base sequence between NSRE-1 and -2 is obligatory in length and highly conserved, and presumably acts as a spacer for one or more molecules linking transcription factors bound to both these elements. The whole complex of bound factors is necessary for transcription (Zhong et al., 2003), and hence a protein binding to the linker could additionally interfere with complex formation.

Figure 7
figure7

Binding site of ASMB in relation to the ASY response unit

The relative significance of this mechanism to the overall repression of ASY has not been determined. It is possible that ASMB forms a repression complex that brings about modification of histones related to the ASY promoter as discussed above in the case of other methyl binding proteins. The recent finding that Kaiso, which binds to specific DNA sequences, also acts in this way through the co-repressor protein N-CoR (Yoon et al., 2003) directs further attention to such a possibility.

Finally, since ASMB protein is widely distributed, at least in lymphoid cells, it seems unlikely that the specific binding sequence now observed has methylated ASY as its only site of action. With these questions in mind, the unique characteristics of the ASMB protein make it deserving of continuing study.

Materials and methods

Leukemic cell lines

Cell lines were obtained originally from the American Type Culture Collection. 1873 (RS4;11) is probably of dual lineage with B-cell characteristics. The wild-type cells are ASN auxotrophs. An ASN-independent variant (1873-IND) derived from the wild type by culture for 50 days in an ASN-depleted medium shows no change in karyotype on Giemsa banding. 1929 (SUP-B15), also an ASN auxotroph, is derived from the marrow of acute lymphoblastic leukemia, and expresses B lineage markers. In one set of experiments, 1929 and 1873 cells were co-cultured for 1 month or more on a layer of mouse macrophages in an ASN-depleted medium. The leukemic cells were readily suspended, while macrophages remained firmly attached to the surface of the flask. The human cells showed complete demethylation of ASY promoters. The REH cell line (ASN prototroph) has cellular characteristics of early B-ALL and shows TEL-AML1 gene fusion (Matheson and Hall, 2003). The characteristics of these cell lines, and also Raji, Jurkat and NB2 are found in Drexler (2001).

Proliferation assay for leukemic cells

Proliferation assays were performed by using a CellTiter 96® Nonradioactive Cell Proliferation Assay kit (Promega) according to the manufacturer's instructions. In total, 100 μl of DMEM containing various concentrations of ASN with 4 × 104 cells of individual leukemic cell lines was added to each well of the 96-well plate. Samples were set up in triplicate. Cell-free media with and without ASN were added as controls.

DNA extraction and methylation analysis

Genomic DNA was extracted by DNAzol® Reagent according to the instructions of the manufacturer (Invitrogen). Methylation analysis was described previously (Peng et al., 2001). After bisulfite conversion, specific PCR amplification was performed using primers that consisted of XLS012 (5′-IndexTermIndexTermCTA AAA CCA AAA ATA TAA ACA AC-3′) and XLS016 (5′-IndexTermIndexTermTTA GAA TAG TAG GTA GTT TGG G-3′). The PCR product was cloned into plasmid PCRR2.1 vector using the original TA cloning kit (Invitrogen). Clones that contained inserts of correct size, as determined in 8% polyacrylamide gels with TBE buffer, were sequenced.

mRNA expression analysis by RT–PCR and TaqMan real-time PCR

RNA was extracted from centrifuged leukemic cell pellets using Trizol® reagent (Invitrogen). Complementary DNA was synthesized from 5 μg of total RNA using the Advantage™ RT-for-PCR kit (Clontech) following the manufacturer's instructions with a 40 μl volume for the RT reaction. PCR reaction was performed in 50 μl volumes consisting of 1.0 μl each of 10 mM of mixed dNTPs, 4.0 μl of 25 mM MgCl2, 1.25 U of AmpliTaq™ Gold DNA polymerase (Applied Biosystems), 2.0 μl of cDNA reaction mixture and 5.0 μl of 10 × PCR buffer. The primers for amplification of human ASY cDNA were derived from the Gene Bank Accession # NM-133436, as forward primer 5′-IndexTermIndexTermTCA GAA CAG CAG GTA GCC TGG-3′ and reverse primer 5′-IndexTermIndexTermCTG AGG CCA GGG ATG TGG ACA-3′. Significantly for co-cultivation experiments, the mouse ASY equivalent was not amplified with the primers. Human GAPDH mRNA was used as an internal control. After preliminary, nonquantitative assay, real-time PCR was performed in a model 7700 sequence detector (PE, Applied Biosystems) using a TaqMan™ PCR kit. Briefly, PCR was conducted under conditions of 10 min at 95°C, 40 cycles of 20 s at 95°C and 1 min at 62°C in 25 μl of reaction. This contains 2.5 μl of 10 × TaqMan™ buffer, 5.5 μl of 25 mM MgCl2, 0.5 μl of each 10 mM dCTP; dATP; dGTP; dUTP, 1.25 μl of 2 μ M FAM labeled probe (5′-IndexTermIndexTermCGC GTG GTC CTG ATC TAG GAA GAG ACT G-3′), 3.125 μl each of 2 μ M Primers (Forward primer: 5′-IndexTermIndexTermCCA GGG CGC AGT CAG GT-3′; Reverse primer: 5′-IndexTermIndexTermACC GAG ACG GTG ATAAGG CC-3′). Real-time fluorescence measurements were taken and the threshold cycle (CT) value for each sample was calculated by determining the point at which the fluorescence exceeds a threshold limit. At the same time, the primers and probe of GAPDH were added in the same reaction for normalization of RNA amount of each sample. The relative quantitation of ASNS expression was calculated by the formula 2−ΔΔCt.

Oligonucleotide labeling

Oligonucleotide probes derived from the promoting region of ASY had the following sequences: distal (−151 to –100) 5′ IndexTermIndexTermCTTCCGCCGCCCCACATTAGTCCTGCTCCGCCCCGGACACCCC. Middle (−105 to –71) 5′ IndexTermIndexTermCCCCGCGGCCCCGCCCCTGTGCGCGCTGGTTCCTCCTCGCAG. Proximal (−75 to –33) 5′ IndexTermIndexTermGCAGGCATGATGAAACTTCCCGCACGCGTTACAGGAGCCAGG. Oligonucleotides and their complementary sequences (including methylated forms and variants) were synthetized by Integrated DNA Technologies Inc.

Double-stranded oligonucleotides (200 ng) were labeled with [γ-32P] ATP (3000 Ci/mmol) and T4 polynucleotide kinase (2–3 U/μl, Amersham Bioscience) in a reaction mixture of 50 μl at 37°C for 15 min followed by inactivation of enzyme by incubation at 65°C for 10 min. The labeled oligonucleotides were purified in a Sephadex G-50 spin column (Boehringer). After measuring the radioactivity, oligonucleotides were diluted to 40 000 cpm/μl with 10 mM Tris HCl, pH 7.5, 0.1 M NaCl, 1 mM EDTA and kept at −20°C until use.

Preparation of nuclear extracts

Nuclear extract from leukemic cells was prepared according to the method of Dignam et al. (1983) with some modifications. Cells were washed twice with ice-cold PBS, before being suspended in 5 vol of hypotonic buffer (10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.2 mM PMSF, 1.0 mM benzamidine·HCl, 1% by volume of protease inhibitor cocktail (Sigma) and 14 mM β-mercaptoethanol) and kept on ice for 30 min. Cells were disrupted in a glass homogenizer by 20 strokes using a teflon pestle. The nuclear pellet after recentrifugation was washed twice with hypotonic buffer and suspended in 2 vol of high salt buffer (20 mM HEPES, pH 7.9, 1.5 mM MgCl2, 0.75 M KCl, 14 mM β-mercaptoethanol, 25% glycerol and protease inhibitors). The product was vortexed immediately and kept on ice for 15 min. The final extract was collected by centrifuging at 14 000 r.p.m. for 45 min. The protein content in the extract was measured, aliquoted (50 μl) and kept at −70°C.

Electromobility gel shift assay (EMSA)

Nuclear extract (0.5–1 μg protein) was mixed with 10–20 μg/ml poly(dA/dT), labeled oligonucleotide (40 000 cpm/ng) in a 20 μl reaction mixture containing 25 mM HEPES, pH 7.5, 50 mM NaCl, 3 mM MgCl2, 25 mM KCl, 1 mM DTT, 5% glycerol and incubated at room temperature for 30 min. In competition experiments, 100-fold molar excess of cold oligonucleotide was added. For supershift assays. 2 μg of appropriate antibody was added. The reaction mixture was electrophoresed on 6% polyacrylamide gel in Trisglycine buffer at <4°C. The gel was dried and exposed on Kodak XAR film.

Chromatin immunoprecipitation assay (CHIP)

The method was modified from Boyd et al. (1998). Chromatin proteins of leukemic cells were crosslinked to DNA by addition of formaldehyde directly to the culture medium to a final concentration of 1%. Following a 10-min incubation at room temperature, Glycine was added to a concentration of 0.125 M, and after 5 min cells were washed twice in PBS, and then resuspended in 1% SDS lysis buffer (Upstate Biotechnology). They were sonicated to reduce the size of DNA to 300–1000 bp, as shown by electrophoresis in 10% agarose, and the remaining procedures were performed with reagents (Upstate Biotechnology), according to the manufacturer's instructions. Samples containing 100 μg DNA were used for each assay. A sample of sonicate was saved as ‘input DNA’. Immunoprecipitation used 5 ml anti-ATF4 antibody (Santa Cruz Biotechnology) with overnight incubation at 4°C; a no-antibody control was also performed for each ChIP assay. After extraction with protein A-agarose and washing, the complex was eluted and crosslinks were reversed at 65°C, followed by treatment with ribonuclease and proteinase K. Final samples of DNA were extracted by phenol/chloroform and precipitated with ethanol and glycogen carrier. After dissolving in 40 μl water, amplification was achieved using hot-started reactions with AmpliTaq DNA polymerase (Applied Biosystems) and 5 μl of either immunoprecipitated DNA, a no-antibody control, or a 1 : 100 dilution of input chromatin. The conditions for all reactions were as follows: 95°C for 10 min, 28 cycles at 95°C for 30 s, 62°C for 20 s, 72°C for 30 s and 72°C for 20 min. The primers used are as follows: ASNus: 5′-IndexTermIndexTermTGA TTT CCC GAA GAA ACC AAG TTC-3′; ASNds, 5′-IndexTermIndexTermTGC GGG AAG TTT CAT CAT GC-3′; albumin controls employed: ALBus, 5′-IndexTermIndexTermATT GAC AAG GTC TTG TGG AGA AAA CAG-3′, ALBds 5′-IndexTermIndexTermAAG AGA AAA GCT AGG ACA AAC GGA GG-3′. PCR products were electrophoresed on 2% agarose gels and strained with ethidium-bromide. These conditions were shown to give linear values of amplified DNA over a dilution range of 1 : 50 to 1 : 400 of input DNA Quantitation was by densitometry in a Molecular Imager® FX(Bio-Rad).

Purification of methyl binding protein

In all, 20 μg of ds R15 oligonucleotide (Table 1) was coupled with 0.1 g of CNBR Sepharose 4B (Pharmacia) in 2 ml of buffer (0.1 M NaHCO3, pH 8.3, and 0.5 M NaCl) by incubating at 25°C for 1 h with gentle shaking. The oligomer-bound Sepharose 4B was washed twice with 5 vol of coupling buffer. Any remaining active group in R15-Sepharose 4B was blocked by incubating it in 5 vol of 0.1 M Tris HCl, pH 8.0 at 25°C for 2 h. The gel matrix was washed in three cycles of alternating pH, consisting of 0.1 M Na-acetate, pH 4.0, 0.5 M NaCl followed by 0.1 M Tris.HCl, pH 8.0, 0.5 M NaCl. Finally, R15 conjugated Sepharose was suspended in1 ml of 0.1 M Tris HCl, pH 7.4, 5 mM EDTA, 5 mM ɛ-amino caproic acid.

In all, 100 μl of packed R15-conjugated Sepharose 4B was mixed with 250 μl of nuclear extract (125 μg) in 1 ml of binding buffer (25 mM HEPES, pH 7.5, 50 mM NaCl, 3 mM MgCl2, 25 mM KCl, 20 μg/ml poly(dA/dT), 1 mM DTT, 5% glycerol) and incubated at 25°C for 1 h with gentle shaking. Unbound protein was separated by centrifuging at 1000 g for 1 min. The sepharose conjugate with bound protein was washed five times with binding buffer and once with 100 μl of the same buffer (without poly dA/dT) containing 5 μg of the unmethylated oligomer R10. The methyl binding protein was finally eluted with 100 μl of buffer containing 5 μg of methylated oligomer R15 by gentle vortexing for 30 s at 25° C. Eluates from each stage of washing were electrophoresed in 10% SDS–PAGE and protein was detected by silver staining.

References

  1. Andrulis IL, Chen J and Ray PN . (1987). Mol. Cell Biol., 7, 2435–2443.

  2. Balestar E, Paz MF, Valle L, Wei S, Fraga MF, Espada J, Cigudosa JC, Huang TH and Esteller M . (2003). EMBO J., 22, 6335–6345.

  3. Bird A . (2002). Genes Dev., 16, 6–21.

  4. Bird AP and Wolffe AP . (1999). Cell, 99, 451–454.

  5. Boyd KE, Wells J, Bartley SM and Farnham PJ . (1998). Proc. Natl. Acad. Sci., 95, 13887–13892.

  6. Broome JD . (1963). J. Exp. Med., 118, 99–148.

  7. Broome JD . (1968). J. Exp. Med., 127, 1055–1072.

  8. Daniel JM, Spring CM, Crawford HC, Reynolds AB and Baig A . (2002). Nucleic Acids Res., 30, 2911–2919.

  9. Dignam JD, Lebovitz RM and Roeder RG . (1983). Nucleic Acids Res., 5, 1475–1489.

  10. Drexler HG . (2001). The Leukemia-Lymphoma Cell Line Facts Book. Academic Press: New York.

  11. Ehrlich M and Ehrlich KC . (1993). DNA Methylation: Molecular Biology and Biological Significance. Jost JP and Saluz HP (eds). Birkhauser Verlag: Basel, pp 145–168.

  12. Feng Q and Zhang Y . (2001). Genes Dev., 15, 827–832.

  13. Fuks F, Hurd PJ, Deplos RH and Kouzarides T . (2003). Nucleic Acids Res., 31, 2305–2312.

  14. Greco A, Gong SS, Ittmann M and Basilico C . (1989). Mol. Cell Biol., 9, 2350–2359.

  15. Guerrini L, Gong SS, Mangasarian K and Basilico C . (1993). Mol. Cell Biol., 13, 3202–3212.

  16. Hendrich B and Bird A . (1998). Mol. Cell Biol., 18, 6538–6547.

  17. Holler M, Westin G, Jiricny J and Schaffner W . (1988). Genes Dev., 2, 1127–1135.

  18. Jones PA . (2002). Oncogene, 21, 5358–5360.

  19. Jones PL, Veenstva GJ, Wade PA, Vermaak D, Landsberger N, Strouboulis J and Wolfe AP . (1998). Nat. Genet., 19, 187–191.

  20. Jost JP and Hofsteenge J . (1992). Proc. Natl. Acad. Sci., 89, 9499–9503.

  21. Jost JP and Saluz HP . (1993). DNA Methylation: Molecular Biology and Biological Significance, Jost JP and Saluz HP (eds). Birkhauser Verlag: Basel, pp 425–451.

  22. Khan R, Zhang XY, Supakar PC, Ehrlich KC and Ehrlich M . (1988). J. Biol. Chem., 263, 14374–14383.

  23. Matheson EC and Hall AG . (2003). Carcinogenesis, 24, 31–38.

  24. Meehan RR, Lewis JD, McKay S, Kleiner EL and Bird AP . (1989). Cell, 58, 499–507.

  25. Nan X, Meehan RR and Bird A . (1993). Nucleic Acids Res., 21, 4886–4892.

  26. Ohtani-Fujita N, Fujita T, Aoike A, Osifchin NE, Robbins PD and Sakai T . (1993). Oncogene, 8, 1063–1067.

  27. Peng H, Shen N, Qian L, Sun XL, Koduru P, Godwin LO, Issa JP and Broome JD . (2001). Br. J. Cancer, 85, 930–935.

  28. Ramji DP and Foka P . (2002). Biochem. J., 365 (Part 3), 561–575.

  29. Sengupta PK, Ehrlich M and Smith BD . (1999). J. Biol. Chem., 274, 36649–36655.

  30. Siu F, Bain PJ, LeBlanc-Chaffin R, Chen H and Kilberg MS . (2002). J. Biol. Chem., 277, 24120–24127.

  31. Siu F, Chen C, Zhong C and Kilberg MS . (2001). J. Biol. Chem., 276, 48100–48107.

  32. Wade PA . (2001). Oncogene, 20, 3166–3173.

  33. Yoon HG, Chan DW, Reynolds AB, Qin J and Wong J . (2003). Mol. Cell, 12, 723–734.

  34. Zhong XY, Jabrane-Ferrat N, Asiedu CK, Samac S, Peterlin BM and Ehrlich M . (1993). Mol. Cell. Biol., 13, 6810–6818.

  35. Zhong C, Chen C and Kilberg MS . (2003). Biochem. J., 372 (Part 2), 603–609.

Download references

Acknowledgements

The secretarial assistance of Ms Edda Morgan-Boncy is gratefully acknowledged.

Author information

Correspondence to J D Broome.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • methyl binding protein
  • asparagine synthetase
  • leukemia

Further reading