Interleukin 6 Supports the Maintenance of p53 Tumor Suppressor Gene Promoter Methylation

Epigenetic alteration of DNA represents an important aspect of cell function and is responsible for significant changes that can lead to tumorigenesis (1). The mammalian genome contains only 20% of the predicted CpG dinucleotides with only 70% of this subset being found in a methylated state (2). However, CpG islands found associated with the promoters of many genes maintain the expected frequency of this dinucleotide and remain unmethylated (3). Methylation of the cytosines found in CpG dinucleotides by DNA methyltransferases regulates transcription, ultimately affecting chromatin structure (4). This epigenetic process also serves to reduce the spread of retrotransposons and stabilizes the repetitive DNA elements in the genome (5). Finally, DNA methylation is highly specific throughout the mammalian genome, such specificity being important for embryonic development and proper differentiation of cells (6). Inappropriate methylation of CpG islands is associated with histone deacetylation and gene silencing, and methylation of CpGs outside of CpG islands is associated with significantly higher mutation rates (7, 8).

Hypermethylation of CpG islands located in the promoter regions of tumor suppressor genes has been described in almost all tumors (9). Many other important cellular functions are also affected by this aberrant methyltransferase activity; for example, DNA repair enzymes, hMLH1, MGMT (10); cell cycle checkpoint genes such as Rb (11), p16^INK4a (12), p15^INK4b (13), and p14^ARF (14); apoptosis-related genes such as DAPK (15); and genes involved in the removal and attenuation of oxidative-free radicals, MnSOD (16) and GSTP-1 (17), are all inactivated by promoter methylation. A likely cause of this inactivation is the binding of 5-methylcytosine specific proteins to CpG sites, which interferes with the binding of transcription factors necessary for gene expression (18). However, little is known regarding the signaling mechanisms that drive aberrant methylation and why only certain classes of genes are susceptible to epigenetic silencing.

The use of DNA methyltransferase inhibitor drugs to reduce or block the ability of the DNA cytosine methyltransferases (DNMT) from methylating CpG islands allows the reexpression of previously silenced genes (19). Once reactivated by methyltransferase inhibitors, these formerly silenced genes then carry out their functions of cell cycle arrest, tumor suppression, DNA repair, etc., all functions blocked previously by methylation (19). A problem encountered frequently with nucleotide analogue–based DNA methyltransferase inhibitors is toxicity, as they covalently bind DNA cytosine methyltransferases following incorporation of the drug into DNA (20, 21). It has been suggested that these drugs mimic DNA damage in neoplastic cells, triggering an apoptotic response (21). However, these drugs suffer reduced efficacy due to their incorporation into the DNA of normal cells, resulting in undesired toxic and apoptotic effects (21).

Interleukin 6 (IL6) is an inflammatory cytokine with the ability to induce tumor growth, metastasis, and resistance to chemotherapy in a variety of tumor cells (22). IL-6 is known to activate several signaling pathways, including the Janus-activated kinase/signal transducer and activator of transcription 3 (JAK-STAT3), phosphoinositol-3 kinase (PI-3K), and mitogen-activated protein kinase pathways (22–24). Multiple myeloma is a progressive, neoplastic disease characterized by the expansion of B-lymphocyte clones, during progression to fully differentiated plasma cells (25). Almost all myeloma cells are responsive to IL-6, with many requiring it for growth (22). Furthermore, IL-6 acts as an antiapoptotic mediator in myeloma cells, whereas in normal cells, its functions seem to involve mainly cell differentiation and development (24). Previously, we have described the ability of IL-6 to induce the expression and activity of DNMT-1, the maintenance methylase, in the human erythroleukemia cell line K-562 (26). This finding suggested that IL-6 may affect epigenetic mechanisms outside its previously described functions involving the mediation of inflammation and differentiation. For example, in K562 cells, IL-6 induces the expression of the Fli-1/ERG-B transcription factor by direct activation of STAT3, a transcription factor shown previously to possess oncogenic potential (27). The activation of STAT3 has been shown important in the oncogenesis of many types of cancer (28).

Using the IL-6–responsive multiple myeloma KAS 6/1 and IL-6 autocrine IM-9 cell lines, we show that p53 is epigenetically silenced, resulting in decreased levels of expression. The DNA methyltransferase inhibitor zebularine reversed the methylation present in the promoter of p53, allowing a marked increase in gene expression (29). In addition, when compared with 5-aza-2′-deoxycytidine (decitabine), zebularine seemed less toxic as evidenced by its ability to derepress gene expression in tumor cells without concomitant normal cell death (30). Following drug treatment, the methylation status of p53 seemed a function of whether or not IL-6 was present, because in KAS 6/1 cells the withdrawal of zebularine alone was insufficient for the reestablishment of the original methylation patterns and required the addition of IL-6. This IL-6–mediated remethylation takes place within 48 to 72 hours following the removal of zebularine. Cells that did not receive IL-6 eventually died, presumably because p53 expression continued unabated by remethylation. Further evidence of the importance of IL-6 in maintaining the methylation of p53, comes from using the IL-6 autocrine cell line IM-9. When zebularine was withdrawn from IM-9 cells, the p53 promoter remethylated and the cells survived, without the addition of IL-6. We also show that IL-6 induced the expression of DNMT-1 in KAS 6/1 cells, an observation consistent with our reports in other hematopoietic cell types (26). Our data suggests that the inflammatory cytokine IL-6 is crucial to the establishment and maintenance of p53 promoter methylation, providing a model demonstrating how mediators of inflammation contribute to epigenetic silencing of tumor suppressor genes and tumor cell survival.

Quantitative determination of the ratio of 5-methyl-2′-deoxycytidine to total deoxycytidine residues in genomic DNA by microcolumn liquid chromatography. These studies were carried out using an Agilent Technologies series 1100 Capillary Liquid Chromatograph equipped with a binary pump, online degasser, diode array UV detector with 500 nL flow cell, thermostated column compartment, and thermostated micro autosampler. The mobile phase was composed of 50 μmol/L NaH₂PO₄ (pH 5.0) and methanol delivered at 15 μl/min. A gradient from 2% methanol for 2 minutes, increasing to 8% over the next 13 minutes, and increasing to 60% over the final 10 minutes of a total runtime of 25 minutes was employed. The detector was set to monitor 280 nm to provide maximum sensitivity for 5-methyl-2′-deoxycytidine and peak apex spectra were taken to confirm peak identity. Separation of free deoxyribonucleosides was accomplished using a 150 × 0.5 mm ID microcolumn packed with 5 um Luna C18(2) stationary phase obtained from Phenomenex (Torrance, CA) thermostated at 35°C. Nuclease P1 and bacterial alkaline phosphatase enzymatic digests, containing ∼10 μg of genomic DNA, were dried using vacuum centrifugation and reconstituted in 20 μL of deionized water to yield a concentration of about 0.5 μg/μL. The micro autosampler injected 2 μL onto the analytic column, which is the equivalent of 1 μg of DNA. Analytic reference standards of 5-methyl-2′-deoxycytidine (m⁵dC) and 2′-deoxycytidine (dC) were obtained from Sigma (St. Louis, MO) and stock standard solutions were prepared in deionized water at a concentration of 2 mmol/L each. Working standard mixtures were prepared at the following concentrations of 2′-deoxycytidine: 1,000, 500, 200, 100, 50, and 25 μmol/L with corresponding concentrations for 5-methyl-2′-deoxycytidine of 50, 25, 10, 5, 2.5, and 1.25 μmol/L. Linear calibration curves (R² > 0.9995) were constructed for 2′-deoxycytidine and 5-methyl-2′-deoxycytidine and used to determine nucleoside concentrations in the DNA digests. The molar ratio as a percent was calculated as follows;

The mole percent of 5-methyl-2′-deoxycytidine could be determined in as little as 0.25 μg of total DNA with an absolute error of less than ±0.1% (31).

Cell lines and methyltransferase inhibitor drug treatment. The human multiple myeloma cell lines KAS 6/1 and IM-9 were grown in RPMI 1640 supplemented with 4 mmol/L glutamine, 10% heat-inactivated fetal bovine serum (FBS), and 100 units each of streptomycin and penicillin. KAS 6/1 cells were supplemented with IL-6 at a concentration of 10 ng/mL. Cells were treated with zebularine at a concentration of 250 μmol/L for 48 hours. An analogous 5 μmol/L dose of 5-aza-2′-deoxycytidine (decitabine) was used in one experiment. For the remethylation experiments, cells were washed twice with PBS followed by growth for ∼120 hours in fresh medium with or without IL-6 supplementation.

Genomic DNA isolation and chemical modification. Genomic DNA was isolated from either KAS 6/1 or IM-9 cells using the QIAGEN DNA isolation kit. Genomic DNA was digested with a restriction endonuclease, such as EcoRI, before chemical modification. Digested DNA (4 μg/40 μL) was denatured with 0.3 mol/L NaOH for 15 minutes at 37°C. The denatured DNA was subsequently modified by 3.1 mol/L sodium bisulfite and 0.5 mmol/L hydroquinone in 480 μL under mineral oil for 16 hours at 55°C. Modified DNA was purified using the QIAquick DNA extraction kit, according to the manufacturer's instructions (Qiagen Inc., Valencia, CA). The DNA was eluted with 50 μL of elution buffer (10 mmol/L Tris, pH 8.0). The modification was completed with 0.3 mol/L NaOH treatment for 15 minutes at 37°C. The modified DNA was precipitated by adding 3 mol/L ammonium acetate (pH 7.0) and two volumes of ethanol. The precipitated DNA was resuspended in 50 μL of water for PCR amplification.

PCR of bisulfite-treated DNA. Two sequential PCRs were used to amplify the modified DNA fragments of interest. Two microliters of modified DNA template were used in the first PCR. Nested primers were used in the subsequent PCR. The primers for the p53 gene promoter region are (Genbank accession no. X54156) are (a) first PCR, 5′-TGCCCTCACAGCTCTGGCTTGCAGAATTT-3′ (sense, nucleotides 584-612) and 5′-ACTGAACTTGATGAGTCCTCTCTGAGTCA-3′ (antisense, nucleotides 961-989); (b) nested PCR, 5′-GCAGAATTTTCCACCCCAAAATGTTAGTATC-3′ (sense, nucleotides 604-634) and 5′-CTGGATTGGGTAAGCTCCT-GACTGAACTTG-3′ (antisense, nucleotides 941-970). PCR conditions were as follows: 95C for 30 seconds followed by 35 cycles of 95°C for 30 seconds, 55°C for 30 seconds, 72°C for 2 minutes; and finally 10 minutes at 72°C. The PCR mixture contained 1× buffer (Invitrogen, Carlsbad, CA) with 2.5 units Taq polymerase, 3 mmol/L MgCl₂, 1 μmol/L of each primer, and 0.2 mmol/L deoxynucleotide triphosphates. The amplified DNA fragment of expected size was subjected to cloning, using the pCR4-TOPO TA cloning vector (Invitrogen). Ten individual recombinant clones were sequenced for each group.

Analysis of gene expression by quantitative PCR. Total RNA was extracted from cells using TriZol Reagent (Life Technologies, Inc., Gaithersburg, MD), and cDNAs were prepared by reverse transcribing 1 μg of total RNA using the Single-Strand cDNA Synthesis Kit (Roche Diagnostics Co., Indianapolis, IN) according to manufacturer's protocol. Quantitative PCR (Q-PCR) analysis was done using Taqman probes (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions, in 10 μL final volumes, in 384-well microtiter plates (32). Thermocycling conditions using an Applied Biosystems ABI-7900 SDS were as follows: 50°C for 2 minutes, 95°C for 10 minutes, and 40 cycles of 95°C for 10 seconds and 60°C for 1 minute. Specific primers for Q-PCR of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), p53, and DNMT-1 were designed using Applied Biosystems Assay-by-Design primer design software, and their sequences are proprietary in nature. The target mRNA expression was normalized to the GAPDH expression, and the relative expression was calculated back to the untreated controls for each cell type.

Retrovirus constructs for short hairpin RNA knockdown and expression of WT-p53. Phoenix-Amphotrophic Retroviral Packaging cells were plated at 1.5 to 2 million cells per 60-mm plate in DMEM with 10% FBS, 1% penicillin-streptomycin, 1% glutamine, 18 to 24 hours before transfection. Approximately 2 μg of each plasmid containing the desired inserts (pLZRS-IRES-p53 or pRSMX) were prepared for transfection into cells by using FuGene-6 Transfection Reagent (Roche Diagnostics) according to the manufacturer's instructions. After ∼16 hours, the recombinant retrovirus containing media was filtered through a 0.22-μm disposable filter. Infectious supernatants were supplemented with 5 μg/mL Polybrene to assist in virus attachment. Virus at a multiplicity of infection of ∼10 particles per cell ratio was allowed to infect myeloma cells. Sequences for pRSMX inserts used for p53 knockdown were as follows: (sense) 5′-GATCCCCGACTCCAGTGGTAATCTACTTCAAGAGAG-TAGATTACCACTGGAGTCTTTTTGGAAA-3′ and (antisense) 5′-AGCTTTTCCAAAAAGACTCCAGTGGT-AATCTACTCTCTTGAAGTAGATTACCACTGGAGTCGGG-3′ for insert one; (sense) 5′-GATCCCCGTCTGT-GACTTGCACGTACTTCAAGAGAGTACGTGCAAGTCACAGACTTTTTGGAAC-3′ and (antisense) 5′-AGCTGTTCCAAAAAGTCTGTGACTTGCACGTACTCTCTTGAAGTACGTGCAAGTCACAGACGGG-3′, for insert two.

Ribonuclease protection assays (RNA). Total RNA from untreated, zebularine treated, and cells allowed to reestablish their methylation patterns following withdrawal of zebularine, were extracted by TriZol Reagent and quantified by spectrophotometer. RNase protection assays (RPA) were done according to the manufacturer's protocol (BD Biosciences, San Jose, CA) (RiboQuant RPA Kit). Briefly, 20 μg of total RNA were hybridized overnight at 56°C to 2 × 10⁶ cpm of ³³P-labeled probe corresponding to the multiprobe template set hCC-2, Human Cell Cycle Regulators (BD Biosciences). Unhybridized RNA was digested first with RNase T1 and RNase A for 45 minutes at 30°C and with Proteinase K for 15 minutes at 37°C. After phenol/chloroform extraction and sodium acetate/ethanol precipitation, hybridized RNA probes were denatured at 90°C for 3 minutes and electrophoresed on a 6% polyacrylamide gel. The dried gels were exposed to X-ray film.

Zebularine reduces the levels of 5-methyl-2′-deoxycytidine in KAS 6/1 multiple myeloma cells. Recently, a new DNA methyltransferase inhibitor drug, zebularine, was shown to have considerably reduced toxicity, whereas retaining the ability to reactivate genes silenced by methylation. Zebularine is markedly more stable under physiologic and pharmaceutical conditions than previously used nucleotide analogues thus making it more attractive for clinical use (33). To show that zebularine functioned as a DNMT inhibitor in multiple myeloma cells, we analyzed the levels of 5-methyl-2′-deoxycytidine versus unmethylated 2′-deoxycytidine, in genomic DNA. The myeloma cells were treated with 250 μmol/L zebularine for 48 hours in the presence of 10 ng/mL IL-6 in complete RPMI 1640. DNA from control and zebularine-treated cells was analyzed to determine the ratio of 5-methyl-2′-deoxycytidine to total deoxycytidine by high-performance liquid chromatography (HPLC). As seen in Fig. 1 and the accompanying table, control and remethylated samples are virtually identical with respect to the 5-methyl-2′-deoxycytidine peak, but over one half of the 5-methyl-2′-deoxycytidine was converted to the unmethylated form after treatment with zebularine for 48 hours. Remethylated cells were allowed to recover methylation over ∼1 week's time, their HPLC trace for 5-methyl-2′-deoxycytidine was identical to the control cells (data not shown).

Comparison of effects of decitabine and zebularine on KAS 6/1 multiple myeloma cells. To compare the ability of decitabine versus zebularine to reverse epigenetic silencing of cell cycle and tumor suppressor genes, RPA using total RNAs harvested from KAS 6/1 control cells and cells treated with either 5 μmol/L decitabine or 250 μmol/L zebularine for 48 hours were done. These dose levels represent median levels of the inhibitors and were chosen to provide a side-by-side comparison of their performance. Previously, we determined that decitabine had a very narrow range of efficacy in both myeloma cell lines we treated, doses of 1 μmol/L produced only marginally reduced levels of 5-methyl-2′-deoxycytidine and induced toxic effects if allowed to remained longer than 48 to 72 hours in culture. A dose of 5 μmol/L decitabine produced slightly more reduction in 5-methyl-2′-deoxycytidine, but once again, problems with toxicity were encountered. As shown in Fig. 2A, zebularine elevated p53 expression levels compared with control cells, whereas an equivalent dose of decitabine actually reduced the expression of almost all genes in the group, including a slight reduction in control gene expression. Microscopic examination of the drug-treated cells revealed that significant toxic effects (membrane disruptions, cell debris) were encountered in decitabine-treated cells, whereas zebularine produced few outwardly visible effects. Cell death at the 48-hour time point was determined by trypan blue absorption assay and showed ∼30% to 35% of the decitabine-treated cells took up the dye, whereas zebularine treatment resulted in only 5% to 7% absorption. By comparison, the untreated control cells showed trypan blue absorption levels of about 1% (data not shown). The RPA results and trypan blue absorption assay data indicate that zebularine was superior to decitabine for inducing the expression of important tumor suppressor genes such as p53, presumably repressed by methylation.

Gene resilencing by removal of zebularine from KAS 6/1 and IM-9 multiple myeloma cells. We next compared the expression patterns of the p53 tumor suppressor genes in KAS 6/1 and IM-9 cells. The KAS 6/1 cells are IL-6 responsive multiple myeloma cells, whereas IM-9 is an IL-6 autocrine, multiple myeloma cell line. KAS 6/1 cells grown in complete RPMI 1640 supplemented with 10 ng/mL IL-6 and IM-9 cells grown in complete RPMI 1640 without IL-6 added were both treated with 250 μmol/L zebularine for 48 hours. Following drug treatment, a portion of the cells was rinsed with PBS to remove the drug and transferred to a new flask to recover their methylation patterns. Next, RPAs were done using total RNAs extracted from control (NEG), drug-treated (ZEB), and remethylated (REMET) KAS 6/1 and IM-9 multiple myeloma cells. In Fig. 2B , lane 1 shows expression levels of cell cycle control and tumor suppressor genes before treatment with 250 μmol/L zebularine. The expression levels of p53 were greatly increased as a result of zebularine treatment in both KAS 6/1 and IM-9 cell lines (lane 2), and following withdrawal of zebularine (lane 3), expression returned to pretreatment levels. This suggests that p53, an important apoptotic mediator, was epigenetically silenced by methylation, and that this silencing could be reversed by zebularine. Interestingly, during the course of these experiments we observed that demethylated KAS 6/1 cells survived the remethylation period only when supplemented with IL-6, whereas the autocrine IL-6 producer IM-9 encountered no difficulties. Therefore, we hypothesized that IL-6 was playing a role in mediating the remethylation process.

Bisulfite sequencing analysis of the p53 promoter region. We next examined the methylation status of the p53 promoter region to directly assess the effects of zebularine. Genomic DNA from KAS 6/1 and IM-9 control cells, zebularine-treated cells, and cells allowed to remethylate was chemically modified to convert cytosines into uracils. Methylated cytosines are protected from this modification and thus remain as cytosines when sequenced. We considered a CpG motif to be regulated by methylation if it was methylated in untreated DNA, demethylated when treated with zebularine, and shown to remethylate in a consistent fashion when the drug was removed. The promoter region depicted in Fig. 3, close to the first exon of p53, is shown to undergo both CpG and CpA methylation. A consensus of the methylated sites appears in bold font. The methylation of cytosines found in several CpA sites is a departure from the classic CpG motif, the preferred substrate of DNMT-1. These less commonly seen CpA sites are instead usually modified by the DNMT-3A or DNMT-3B de novo methylases (34). Furthermore, we observed that some of the CpA sites did not remethylate as quickly as the CpG sites, and a few remained unmethylated following drug withdrawal and IL-6 addition. The different recovery kinetics suggest that the methylated CpA sites could be targeted by the de novo DNMT-3A or DNMT-3B methyltransferases and not the DNMT-1 maintenance methyltransferase that remethylates its CpG dinucleotides in a more consistent fashion (35). Based on the observations of the p53 promoter region, it is likely that zebularine not only reverses methylation in CpG sites but also reverses the methylation at CpA sites that are the probable targets of de novo DNMT-3A or DNMT-3B methylases. Also shown in Fig. 3 are the positions of transcription factor binding sites known to be present in the p53 promoter, some of which could be affected by methylated CpG binding proteins.

Remethylation expression time course analysis of p53. To determine the influence of IL-6 on the kinetics of p53 expression in KAS 6/1 and IM-9 cells, a time course of remethylation was done. KAS 6/1 and IM-9 cells were treated with zebularine (day 1) to demethylate the promoters, and the cells were then washed and allowed to recover for 6 days. Levels of expression shown for day 2 represent those following 48 hours treatment with zebularine. We had previously determined this to be the time when remethylation was occurring. In the presence of IL-6, KAS 6/1 p53 expression levels decrease to below pretreatment levels and remain low (Fig. 4A). However, when IL-6 was withheld during the remethylation phase (Fig. 4B), the expression levels of p53 remained at their demethylated levels. Ultimately, these cell growths are arrested and gradually began to undergo apoptosis (data not shown). To assure that this phenomenon was not due to cytokine deprivation induced apoptosis, we also withheld IL-6 from untreated KAS 6/1 cells and observed that the cells did not exhibit any significant increases in apoptosis as seen following drug treatment. Therefore, the observed differences between the results of Fig. 4A and B were not due to simply withholding IL-6. Instead, we suggest a more complex model whereby IL-6 mediates the reestablishment of promoter methylation density that existed before drug treatment. It should be further noted that we determined by sequence analysis that the p53 gene from both KAS 6/1 and IM-9 cells was intact and without mutations (data not shown). Other evidence of intact p53 genes in some multiple myelomas comes from the use of the proteosome inhibitor PS-341 in multiple myeloma cells. PS-341 has previously been shown to induce apoptosis in multiple myeloma by preventing the proteolytic degradation of p53 thus allowing it to function (36). To determine if IL-6 was playing a role in the remethylation process, we extended this analysis to the IM-9 cell line, an autocrine IL-6 producer. Figure 4C shows results very similar to those observed in KAS 6/1 cells following post-drug IL-6 supplementation (Fig. 4A). The IM-9 cells seemed to recover their original methylation levels, as the expression of p53 once again was reduced to slightly below pretreatment levels.

Effects of interleukin 6 stimulation on the expression of the DNMT-1 enzyme. Previously, we reported that IL-6 induced the expression and activity of DNMT-1 in K-562 erythroleukemia cells (26). To determine if a similar mechanism might also be involved in mediating the remethylation phenomenon observed in KAS 6/1 cells, we did real-time PCR analysis. KAS 6/1 cells were rested in low serum (1% FBS) medium for ∼20 hours. The quieted cells were then stimulated with IL-6 at a final cytokine concentration of 100 ng/mL. A fraction of cells were removed before cytokine treatment to serve as a control, and samples were withdrawn for RNA extraction at 2, 4, 6, 8, and 12 hours following IL-6 treatment. A significant increase in DNMT-1 mRNA expression was observed following IL-6 treatment (Fig. 5), beginning at 2 hours post-stimulation, and increasing to a peak level at 6 hours post-stimulation. Thus, IL-6 plays a crucial role in the remethylation process by altering the expression of the DNMT-1 enzyme. Activation of DNMT-1 expression seems responsible for the remethylation of the p53 promoter region.

Role of p53 on apoptosis in KAS 6/1 cells. We next sought to determine the role p53 was playing in KAS 6/1 cells following demethylation and whether or not the induction of apoptosis was indeed mediated by p53. We created three KAS 6/1–based cell lines using the pRSMX retroviral vector, which produces short hairpin RNA products under the control of the H1 promoter. Figure 6A shows the expression levels of p53 protein in each of the three stable cell lines, with an apparent reduction of at least 5-fold expression of p53 in the two knockdown lines. Thus characterized, we next demethylated these three cell lines for 48 hours in 250 μmol/L zebularine, then shifted them into drug-free medium. After 72 hours, the cells were harvested and stained with Annexin V-PE and 7-AAD to determine if reductions in p53 expression correlated with changes in the levels of apoptosis. As shown in Fig. 6B, significantly higher levels of apoptotic cells (lower right quadrant) were observed (47.66%) in the control vector cells compared with the two p53 knockdown cell lines (14.68 and 14.94%, respectively.) The trend was repeated in cells that took up 7-AAD dye and were also labeled with Annexin V-PE (upper right quadrant), representing cells that have lost membrane integrity, probably in late stages of apoptosis. This result shows that following demethylation, the release of p53 from epigenetic silencing seems sufficient to induce apoptosis, which is minimized in both p53 knockdown cell lines. We further examined this system by infecting KAS 6/1 with a retrovirus expressing wild-type p53. Using the actual reverse transcription-PCR generated p53 coding sequence created from KAS 6/1 RNA, we infected cells with intact methylation patterns (no exposure to zebularine) to determine if this was sufficient to induce apoptosis. Approximately 24 hours post-infection, we analyzed the cells using Annexin V-PE and 7-AAD staining to determine levels of apoptosis. Figure 6C shows that retrovirally delivered wild-type p53 induced apoptosis in 74.91% of KAS 6/1 cells compared with only 10.77% in cells transduced with control vector alone thus indicating that supplying p53 in trans is sufficient to induce apoptosis. Furthermore, this result suggests that the other apoptotic mediators are intact in these cells and that methylation-mediated epigenetic silencing of the p53 gene plays a critical role in cell survival.

Analysis of p53 expression in human multiple myeloma cell lines and patient samples. Although mutations within the DNA and tetrameric binding domains of the p53 gene represent a common mode of inactivation, the presence of such mutations in human multiple myeloma is rare (37). Furthermore, in the cell lines used for this report, we found no mutations in the coding regions of any p53 cDNAs we isolated (data not shown). Gene expression profiling has already been done on many multiple myeloma cell lines and patient samples (38, 39). To extend our studies beyond KAS 6/1 and IM-9, we analyzed RNA expression levels in multiple myeloma cell lines profiled by Hurt et al. (39). Figure 7A shows a strong correlation among IL-6, DNMT-1, and p53 expression levels. In other words, high IL-6 expression is concordant with high DNMT-1 expression and low levels of p53, possibly due to methylated p53 promoters. To determine if p53 expression is altered in multiple myeloma patients, we looked at its expression in patient samples profiled by Zhan et al. (40). Figure 7B shows the broad expression range of p53. Over one half of the patients show significantly reduced expression of p53, with the extremes representing an 11-fold difference in expression levels. Moreover, these findings are consistent with the data we obtained with KAS-6/1 and IM-9 suggesting that the epigenetic silencing of p53 is a common feature among myeloma patients.

Recent reports have suggested a connection between inflammation and cancer, showing that many different types of tumor cells proliferate in response to the proinflammatory cytokine IL-6. Whereas the exact nature of the effects of IL-6 on a particular cell type varies depending upon the model, it is well documented that IL-6 can activate at least three major pathways of cell proliferation. Furthermore, the ability of IL-6 to serve as an antiapoptotic factor by activating the PI-3K/AKT pathway also provides myeloma cells with the capability of resisting chemotherapeutic intervention. However, little is known of how inflammation induces or initiates tumors, or if cytokines such as IL-6 serve only to promote the growth of already established tumors. In this report, we have shown an important link between IL-6 signaling and methylation. The epigenetic silencing of important cell cycle, tumor suppressor, and inhibitor genes has been reported previously in multiple myeloma cells (41). The p16^INK4 gene is epigenetically regulated in many tumors, and we observed this in our screening of multiple myeloma cell lines as well (41). In addition, we have reported that the expression of several members of the Bcl-2 family (BAD, BAK, BIK, and BAX) are also affected by methylation-induced silencing (41). Also interesting is the finding that the suppressor of cytokine signaling-1 (SOCS-1), an inhibitor protein that negatively regulates signaling from the IL-6–mediated JAK-STAT pathway, is silenced by virtue of its hypermethylated promoter in some autocrine IL-6 multiple myeloma cells (38–40).

As shown by the microarray data (38–40), the reduced expression of p53 is a common theme among isolates from myeloma patients and cell lines. Over one half of all myeloma cell lines and patients have significantly reduced expression levels of p53. Moreover the coordinated expression of IL-6 and DNMT-1, which inversely correlate to p53 expression levels, suggests that the ability of IL-6 to induce methylation of the p53 promoter through up-regulation of DNMT-1, is a common mechanism of p53 inactivation in multiple myeloma cells (42). Clearly, loss of p53 expression is a critical event in these cells, as shown by our knockdown and reconstitution experiments. The combined action of IL-6 to mediate both promoter methylation and escape from apoptosis (43), provides an attractive model of how chronic inflammation could lead to cancer. As a central mediator in the initiation of apoptosis, p53 is often mutated in tumors, and loss of p53 function is an important prognostic indicator. When examined in the context of our findings, along with the fact that mutations in p53 are indeed rare in this particular malignancy (37), the data strongly suggest that inflammation, in the form of IL-6, could function to gradually alter the epigenetic programming of the developing B lymphocyte, resulting in the reduced expression of p53 seen in many multiple myeloma cells.

Cytosine methylation is an epigenetic modification that alters the normal gene expression of cells, reducing the expression of p53 below functional levels. This loss of p53 expression was reversed by treating the cells with zebularine, a known cytidine deaminase inhibitor recently described to possess DNA methyltransferase inhibitor properties (44). However, the reversal of cytosine methylation is not absolute, as in order for the drug to function it must be incorporated into DNA, where it can covalently bind the methyltransferase enzymes (44). Because DNMT-1 preferentially targets CpG sites containing one methylated cytosine (hemi-methylated), other Cp“X” sites, such as CpA, may be targeted by the de novo DNMT-3A and DNMT-3B methylases (35). By comparing the status of promoter methylation with the expression levels of the p53 mRNA, we were able to show a clear, inverse correlation between expression and increased methylation. Moreover, when the drug was removed, in the presence of IL-6 many of the methylation sites returned to their original state, and repression of p53 mRNA expression occurred. Clearly, zebularine inhibited the actions of the methyltransferases but did so against the background of constant methyltransferase expression and activity, which likely maintains at least some minimal level of methylation of CpG and CpA sites.

The binding of methylcytosine specific proteins to various promoter and enhancer sequences is known to interfere with the binding of transcription factors, especially SP-1, whose DNA binding sites contains a G/C-rich motif (45). Whereas the CpG density of the p53 promoter region is less than what has been observed in other promoter regions, it has been shown that merely one methylated residue was sufficient to repress expression of p53 (46). The ETS family of transcription factors is known to transactivate p53 (47), and several ETS binding sites are present within the promoter region we analyzed. Also in close proximity to these ETS sites are multiple methylated cytosines, which could interfere directly by steric competition for the site, as the “footprint” for the ETS family proteins extends beyond the core binding element (47). The YY-1 transcription factor also transactivates the p53 promoter (48). However, it has been shown that cytosine methylation does not block the DNA binding and function of YY-1, whereas at the same time, is sufficient to prevent ETS factor binding, abolishing its ability to drive expression (49). Other crucial transcription factor binding sites for activator protein (AP-1) and NF-6B are also present, with the AP-1 site found within the region containing the highest density of methylation sites. The close proximity of the methylated cytosine motifs and the transcription factor binding sites adjacent to them, likely leads to the reduced expression of p53 because of methylation induced changes in chromatin structure and accessibility (4).

Previously, we have shown the effects of IL-6 on DNMT-1 regulation in other cell models (26), and our present data once again confirm the ability of IL-6 to induce the expression of DNMT-1 in hematopoietic cells. The requirement for IL-6 in the methylation recovery step broadens the range of effects for this central mediator of inflammation and suggests that this could be one of the reasons tumors, or cells progressing along neoplastic pathways develop a dependence upon IL-6. Moreover, because deprivation of IL-6 resulted in continued expression of p53 and cell death in KAS 6/1 cells and supplementation with IL-6 leads to the remethylation of the p53 promoter, this strongly implicates IL-6 as a contributor to the maintenance of promoter methylation. This observation is further supported by our use of the autocrine IL-6 cell line IM-9, whose p53 gene expression is increased following demethylation but is quickly down-regulated following removal of zebularine.

Here we show that IL-6 facilitates the remethylation patterning and epigenetic gene silencing of p53, an important cell cycle control and tumor suppressor gene, in part by modulating the expression levels of DNA methyltransferase, DNMT-1. The notion that chronic inflammation is an important factor in the development of neoplasia has been shown previously by seminal work on the generation of plasmacytomas in the murine model (50). Moreover, the induction of such tumors in animals in which the IL-6 gene was disrupted, was greatly reduced, suggesting an important role for IL-6 in tumor development (51). Therefore, it is possible that an unintended consequence of increased IL-6 activity may in fact be its ability to induce epigenetic gene silencing by the alteration of DNMT-1 expression patterns, resulting in the disruption of epigenetic programming. Our data shows a clear association between mediators of inflammation and DNA methylation in the epigenetic control of tumor cell functions, and provides one possible pathway regarding the inflammatory mediated initiation of neoplastic growth.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.