Intestinal epithelial cells play critical roles in regulating mucosal immunity. Since epigenetic factors such as DNA methylation and histone modifications are implicated in aging, carcinogenesis, and immunity, we set out to assess any role for epigenetic factors in the regulation of intestinal epithelial cell immune responses. Experiments were conducted using the HCT116 cell line, and a subclone was genetically engineered to lack DNA methyltransferases (DNMT). The induction of the chemokine interleukin-8 and the antiapoptotic protein cFLIP by tumor necrosis factor-α were markedly less in HCT116 cells lacking DNMT than in parental cells. These effects were accompanied by lower monocyte chemotaxis and higher caspase signaling in HCT116 cells lacking DNMT than parental cells. Tumor necrosis factor-α-induced NF-κB activation was blocked and IκBα expression was higher in HCT116 cells lacking DNMT than in parental cells. A CpG island in the IκBα gene promoter region was found to contain variable levels of methylation in parental HCT116 cells. Chromatin immunoprecipitation analysis of histone proteins bound to the IκBα gene promoter revealed that higher levels of IκBα expression in HCT116 cells lacking DNMT compared with parental cells were accompanied by more chromatin marks permissive to gene transcription. These findings show that epigenetic factors influence the NF-κB system in intestinal epithelial cells, resulting in a previously unrecognized mechanism of innate immune regulation.
- mucosal immunity
- DNA methylation
intestinal epithelial cells occupy a critical position at the host-lumen interface of the alimentary tract. Current paradigms of mucosal immunity invoke important roles for intestinal epithelial cells in the regulation of normal immune homeostasis in the gut. These cells are capable of processing and presenting luminal antigens (3, 16); sensing pathogen-associated molecular patterns (1, 6, 40); and highly regulated secretion of soluble mediators of inflammation such as cytokines, chemokines, and eicosanoids (9, 23). Through these processes, intestinal epithelial cells exert strong control over the timing, magnitude, and nature of innate immune responses in the gut following exposure to injurious stimuli and in chronic inflammatory diseases.
The transcription factor NF-κB has been proven to be a dominant regulator of many immune responses of intestinal epithelial cells, through its role in increasing the expression of a program of specific target genes (11). Furthermore, NF-κB plays a critical role in the neoplastic transformation of intestinal epithelial cells (14). Much is already known about the molecular regulation of NF-κB in the setting of acute inflammation. Cytokine- or pathogen-induced activation of NF-κB depends on the degradation of its endogenous inhibitor IκBα, and following acute NF-κB activation the pathway is shut off by resynthesis of IκBα, itself an NF-κB target gene (24). However, little is understood about factors that control basal and inducible levels of activity of this pivotal transcription factor, knowledge that is important for unraveling its role in chronic inflammatory diseases and cancers of the gut.
The methylation of cytosines in CpG dinucleotides and the posttranslational modification of histone proteins in chromatin at gene regulatory regions are two key mechanisms that regulate the expression levels of those genes. These epigenetic factors can change with aging (20) and in the development of neoplasia, in which tumor suppressor gene suppression by methylation of CpG islands can play a critical role in tumorigenesis (34). Recent evidence has shown that chronic inflammation can induce DNA methylation (19), but little is known about how methylation may regulate inflammation.
In this work, we have explored potential roles for epigenetic factors in the control of intestinal epithelial cell inflammatory responses mediated by NF-κB. Our findings led us to characterize the control of the IκBα promoter by epigenetic factors and assess how those factors affect NF-κB activity and target gene expression. The results reveal that the methylation of CpG dinucleotides in the promoter of IκBα exerts strong influences on the NF-κB system and consequently on immune functions of intestinal epithelial cells. This information shows how epigenetic factors can regulate inflammatory responses of intestinal epithelial cells and may therefore play a role in fine tuning intestinal immune homeostasis.
MATERIALS AND METHODS
HCT116 and HCT116 double knockout (DKO) cells were gifts from Prof. Bert Vogelstein at The Johns Hopkins University of Medicine, Baltimore, MD and were cultured in McCoy's 5A medium (Invitrogen, Paisley, UK). THP-1 cells were grown in RPMI 1460 medium (Sigma-Aldrich). Medium was supplemented with 2 mM l-glutamine and 10% heat-inactivated FBS (Sigma-Aldrich).
HCT116 DKO cells were engineered to have defective DNA methylation by deletion of the genes encoding DNA methyltransferase-1 and -3b (29). Prior reports of HCT116 DKO cells indicated greater than 95% reduction in methylated CpG DNA compared with parental cells. mRNA expression analysis of HCT116 DKO cells demonstrated that DNMT1 mRNA levels were in fact comparable to the parental cells but that DNMT3b mRNA expression was abolished (data not shown). Quantitative RT-PCR (qPCR) showed that XAGE-1, an epigenetically silenced gene, is expressed at low levels in HCT116 parental cells but that the DKO cells express severalfold higher levels of this gene (data not shown), showing that these cells have defective epigenetic gene silencing.
Whole-cell or nuclear extracts were prepared, and protein concentration was determined by use of the BCA assay kit (Thermo Scientific, Rockford, IL). Western blotting was done as previously described (33a). Western blotting for TNFR1 was done under nonreducing conditions. Primary antibodies were IκBα (Cell Signaling, Danvers, MA, no. 9242), phosphor-p65 (Ser536) (Cell Signaling, no. 3036s), FLIP (Cell Signaling, no. 3210), caspase 3 (Cell Signaling, no. 9662), TNFR1 (R&D Systems, Minneapolis, MN, no. MAB225), β-actin (Sigma-Aldrich, A5316), and Lamin B1 (Abcam, Cambridge, UK, ab16048). A G:BOX chemi-system (Syngene, Cambridge, UK) was used to capture luminescent signals.
qPCR was performed as previously described (10). Primer sequences were designed using the Primer 3 algorithm (30) and are available on request. GAPDH was used as an internal control, and relative expression levels were calculated by the ΔCT method (28).
ELISA for IL-8.
HCT116 and HCT116 DKO cells were seeded at 0.3 × 106 cells/well of a six-well plate. At 24 h after seeding cells were treated with 10 ng/ml TNF-α, in 1 ml of medium, for 6 h. The medium was removed and used in the ELISA reaction. The ELISA was done by using the Quantikine Human CXCL8/IL-8 ELISA kit from R&D Systems according to the manufacturer's instructions.
Caspase 3 assay.
Media and cells were collected and spun at 1,000 g for 4 min. The cell pellets were washed with 1 ml of cold PBS, and 100 μl was retained for BCA assay. Remaining cells were spun again at 1,000 g for 4 min. Cells were then lysed with 100 μl lysis buffer [HEPES buffer (50 mM HEPES-KOH pH 7.2, 5 mM EGTA, 10 mM KCl, 2 mM MgCl2) + 2 mM DTT and 0.1% 3–[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate (CHAPS)] on ice, and 100 μl of cell lysate for each sample to be analyzed was added to a different well of a 96-well black plate. We added 100 μl of 2× substrate solution [HEPES buffer + 2 mM DTT and 40 μM Ac-DEVD-AFC] immediately to each well. Green fluorescence was read with excitation at 400 nm and emission at 505 nm. The caspase activity was determined by the rate of increase of fluorescence (slope) normalized by the protein content (determined by BCA assay) and expressed as milli-arbitrary fluorescence units per minute per microgram per milliliter.
The chemotaxis assay was carried out by use of the CytoSelect Cell Migration assay kit from Cell Biolabs (San Diego, CA). HCT116 and HCT116 DKO cells were treated with 10 ng/ml TNF-α for 24 h, and 3 × 105 THP-1 monocytes were placed inside the porous insert and conditioned medium from HCT116 or HCT116 DKO cells were placed outside the insert. THP-1 cells were allowed to migrate for 24 h at 37°C. Migrated cells were lysed and detected by CyQuant GR dye (Invitrogen). Fluorescence was measured at 480/520 nm.
p65 DNA binding assay.
p65 DNA binding was analyzed from whole cell extracts using the TransAM NF-κB p65 transcription factor binding assay (Active Motif, Rixensart, Belgium) according to the manufacturer's instructions.
Genomic DNA preparation and bisulfite sequencing.
Genomic DNA was isolated by using the DNeasy Tissue Kit (Qiagen) according to the manufacturer's instructions. DNA was bisulfite treated by using the EpiTect Bisulfite Kit (Qiagen) according to the manufacturer's instructions. Primers were designed with Methyl Primer Express v1.0 (Applied Biosystems) and based on the NFKBIA gene sequence (NM_020529) and promoter sequence from PromoSer (http://biowulf.bu.edu/zlab/PromoSer/). Two sets of primers were designed to amplify the IκBα CpG island. Primer set 1 was designed to recognize the IκBα promoter (CpG I) and primer set 2 was designed to recognize the promoter/exon 1 (CpG II). Primer sequences were designed for the bisulfite converted bottom strand and did not contain any CpG dinucleotides, sequences are available on request. Each 50 μl PCR reaction contained 1 μg of bisulfite-converted gDNA, 1× NH4+ (Bioline, London, UK), 2 mM Mg2+ (Bioline), 5 μl dNTP mix (Bioline), 2 μl (10 pmol/μl) of each primer, 1 U Taq polymerase (Bioline), and DNase/RNase-free H2O. Cycling conditions were as follows: 40 cycles of 95°C for 15 s followed by annealing at 55°C for 30 s and extension at 72°C for 15 s. PCR products were purified by using the QIAquick gel extraction kit (Qiagen) according to the manufacturer's instructions. These PCR products were TA cloned by using the pGEM-T Easy Vector System (Promega, Southampton, UK) according to the manufacturer's instructions. Nine of ten clones were miniprepped by using QIAprep spin Minikit (Qiagen) according to the manufacturer's instructions and were sent to Eurofins/MWG, Ebersberg, Germany, for sequencing.
Combined bisulfite restriction analysis (COBRA) was carried out on PCR products of the CpGII region of the IκBα gene, from bisulfate converted DNA from HCT116 and DKO cells. PCR products were incubated with the BstUI restriction enzyme (New England Biolabs, Ipswich, MA) under the conditions recommended by the manufacturer. BstUI recognizes and cleaves 5′-CG*CG-3′ sequences.
Native chromatin was prepared from ∼2 × 107 HCT116 and HCT116 DKO cells. We used 150 μg of native chromatin for each immunoprecipitation according to the Abcam protocol. Antibodies used were rabbit polyclonal directed against histone H3 (tri methyl K9) (Abcam), histone H3 (acetyl K14) (Millipore), and rabbit monoclonal directed against histone H3 (di methyl K4) as well as an irrelevant control (rabbit IgG). In each chromatin immunoprecipitation (ChIP) reaction 15 μg of antibody was used. DNA was isolated by phenol-chloroform extraction and PCR cleanup columns [QIAquick Gel Extraction Kit (Qiagen)]. This DNA was used as the template in qPCR reactions by using primers specific for the IκBα promoter and first exon. Relative fold enrichment was calculated by the ΔCT method, normalized to irrelevant IgG binding. Primer sequences are available on request.
Data are expressed as means ± SD. Comparison between two groups was done by a Student's t-test. A probability value of 0.05 or less was considered significant.
Lower IL-8 secretion and reduced monocyte chemotaxis by HCT116 DKO than parental cells.
First, we assessed the responses of HCT116 cells to stimulation with TNF-α, a potent proinflammatory cytokine. As assessed by the expression of IL-8 mRNA, HCT116 cells exhibit increased expression of this chemokine almost 40-fold following TNF-α treatment (Fig. 1A). By contrast, DKO cells exhibit increased expression of IL-8 mRNA following stimulation, but to a significantly lesser degree than parental cells. The secretion of IL-8 protein by HCT116 parental cells was also much greater than by DKO cells (Fig. 1B). Next, we assessed the functional effects of DNA methyltransferases (DNMT) on HCT116 cells ability to induce chemotaxis of monocytes. Conditioned medium from TNF-α-stimulated HCT116 parental or DKO cells was added to a culture system containing THP1 monocytic cells in a porous insert. Fewer THP1 cells migrated toward the DKO-conditioned medium than control parental-conditioned medium (Fig. 1C). This finding is consistent with the lower expression of IL-8, and possibly other chemokines in DKO-conditioned medium. To exclude the possibility that the differential responses to TNF-α were simply due to lower abundance of TNF receptors in DKO cells, we compared expression of the TNF receptor-1 in HCT116 parental and DKO cells. Real-time RT-PCR and Western blotting showed very similar levels of TNF receptor-1 in both cell lines, excluding this possibility (Fig. 1D) Together, these findings indicate that DNMT affects the ability of HCT116 cells to induce monocyte chemotaxis, an important proinflammatory event.
HCT116 DKO cells have lower cFLIP expression and higher caspase activity than parental cells.
In addition to the regulation of mucosal immunity through the secretion of soluble factors, intestinal epithelial cells also constitute a barrier between host and lumen. The apoptosis of intestinal epithelial cells partially controls their ability to constitute their barrier function (2). Therefore, we also assessed the effect of DNMT on the responses of HCT116 cells to TNF-α-induced expression of cFLIP, a dominant controller of death receptor-mediated apoptosis. Results revealed that the expression of cFLIP mRNA (Fig. 2A) and protein (Fig. 2B) was similar between parental and DKO HCT116 cells at baseline. However, TNF-α treatment increased expression of cFLIP to a greater extent in the HCT116 parental cells than the DKO cells. This suggests that the cells lacking DNMT might activate apoptotic signaling to a greater extent than parental cells. We assessed this possibility using a fluorogenic caspase-3 activity assay. The results showed that TNF-α treatment resulted in a robust activation of caspase-3 activity in HCT116 cells, which was somewhat greater in DKO than in parental cells (Fig. 2C). Western blotting for caspase 3 levels showed no differences in expression between HCT116 parental cells and DKO cells (Fig. 2D), suggesting that the greater caspase 3 activity in DKO cells is due to lower levels of the caspase inhibitor cFLIP, not differential expression. This role for DNMT in controlling the expression of apoptosis regulating proteins and caspase signaling suggests that intestinal epithelial homeostasis may be under the influence of DNA methylation.
HCT116 DKO cells have lower NF-κB activity than parental cells.
Responses of intestinal epithelial cells to TNF-α are strongly controlled by the transcription factor NF-κB. Moreover, both IL-8 and cFLIP are known NF-κB target genes. Therefore, we reasoned that the differences we observed in IL-8 and chemotaxis and in the cFLIP and caspase activity between HCT116 parental and DKO cells might be related to different levels of NF-κB activity. We used an NF-κB DNA binding assay to compare basal levels of activity of the transcription factor between HCT116 parental and DKO cells. Results revealed that basal NF-κB DNA binding activity is approximately half in DKO compared with parental HCT116 cells (Fig. 3A). Next, we compared the levels of induction of NF-κB activity by TNF-α, a prototypic inducer of this transcription factor. TNF-α significantly increased NF-κB DNA binding activity in parental HCT116 cells but caused no increase in DKO cells (Fig. 3B). The phosphorylation of the NF-κB p65 subunit at certain serine residues is an essential posttranslational modification for NF-κB DNA binding activity (7). Phosphorylation at serine 536 is regulated by IKKβ in response to TNF-α stimulation (39), and mutation of this serine to alanine has been shown to block the binding of p65 to CBP/p300 (8). Immunoblotting for nuclear phosphorylated p65 (Ser536) revealed that TNF-α induced little phosphorylation of p65 in the DKO cells compared with parental cells (Fig. 3C). These findings show that DNMT regulates basal and cytokine-induced NF-κB activation, but the mechanism required additional investigation.
HCT116 DKO cells have higher IκBα expression than parental cells.
In the classical NF-κB activation pathway, stimulation by TNF-α leads to phosphorylation of IκBα and its subsequent degradation by the proteasome. NF-κB is then free to translocate to the nucleus where it can bind to NF-κB sites on the promoters of its target genes, upregulating their expression (13). Since IκBα strongly controls cytokine-induced NF-κB activation, we compared the expression levels of IκBα in HCT116 parental and DKO cells. At the mRNA level, HCT116 DKO cells expressed IκBα at approximately double the level found in parental cells (Fig. 4A). At the protein level, HCT116 DKO cells also expressed approximately two or three times more IκBα under basal conditions (Fig. 4B). Treatment of HCT116 parental cells with 10 ng/ml TNF-α for 30 min resulted in the typical pattern of IκBα protein degradation, which is associated with NF-κB activation. In HCT116 DKO cells, IκBα protein was lower after TNF-α treatment, but did not completely disappear. Densitometry showed that the relative decline in IκBα levels following TNF-α stimulation was similar in the HCT116 and DKO cells, and that there was significant residual IκBα remaining only in DKO cells. Therefore, the finding of significant residual IκBα expression following TNF-α stimulation of HCT116 DKO cells may explain the inability of this cytokine to activate NF-κB. These results indicated a role for DNMT in the suppression of cellular IκBα levels in HCT116 cells, in both basal and stimulated conditions, which appears to correlate with both basal and stimulated NF-κB activity.
The IκBα promoter contains a CpG island that shows heterogeneous DNA methylation in HCT116 cells.
To determine whether DNMT-regulated IkBα expression was mediated by promoter DNA methylation, we examined the IκBα promoter region for CpG islands using Methyl Primer Express v1.0 software. A 1,882-base pair CpG island was found in the promoter, beginning 491 bases upstream of the transcription start site (21) and extending into the second exon (Fig. 5A). The CpG island sequence was taken to the first exon and divided into two sections (Fig. 5B) for bisulfite sequencing. We chose to analyze the bottom strand for bisulfite sequencing because it has been reported that there is a PCR bias in bisulfite sequencing toward the top strand (37). IκBα CpG I was found to be unmethylated in both HCT116 and DKO cells (data not shown). IκBα CpG II displayed high levels of methylation in HCT116 cells in two of ten clones sequenced. In DKO cells IκBα CpG II was found to be largely unmethylated (Fig. 5C). To confirm this finding, COBRA was done on CpG II (bottom strand) bisulfite-converted PCR products by using BstUI restriction enzyme (Fig. 5D). BstUI recognizes and digests DNA that contains 5′CG^CG3′3′ or 3′GC^GC5′ sequence. In unmethylated DNA all cytosines are converted to thymines so there are no recognition sites for the enzyme. There are five recognition sites for this enzyme within the completely methylated CpG II sequence; the number of digested fragments is directly proportional to the degree of methylation. The positive control was completely digested with the small digested fragments migrating off the gel, and the negative control was completely undigested. HCT116 CpG II was partially digested, resulting in two small fragments, and DKO CpG II remained undigested. The results of the COBRA study confirm the bisulfite sequencing data, showing that the IκBα promoter displays significant methylation in HCT116 cells, albeit in a heterogeneous fashion, and is unmethylated in DKO cells. These results suggest that the increased IκBα expression found in the DKO cells is due to decreased DNA methylation at the IκBα promoter and consequently lower levels of gene silencing. IκBα CpG island methylation was quite variable in the different clones we isolated, suggesting heterogeneity of cells in the HCT116 parental line.
Histone modifications associated with active transcription are increased at the IκBα promoter in DKO cells.
Promoter activity is also regulated by complex interactions between DNA methylation and a variety of posttranslational modifications of chromatin. Certain histone modifications are associated with active gene transcription whereas others are linked to gene repression. We used native ChIP to investigate protein-DNA interactions at the IκBα promoter, comparing HCT116 parental and DKO cells. Ac-H3K14 and H3K4(Me2) have been shown to be primarily associated with 5′ regions of transcriptionally active genes and less so downstream of the transcription start site (27). H3K9(Me3) has been reported to be also found at the transcribed regions of active mammalian genes (4, 32, 35, 36). Analysis of ChIP results performed with specific antibodies against these three histone variants revealed that all of them were detected bound to the IκBα promoter and the first exon (Fig. 6). We detected increased levels of Ac-H3K14, H3K4(Me2), and H3K9(Me3) bound to the IκBα promoter in HCT116 DKO cells than in parental cells. Relatively small amounts of these histone variants were detected binding to the first exon of the IκBα gene, with the exception of H3K9(Me3). Similar to the promoter region, higher amounts of these histone marks were detected binding to first exon DNA from HCT116 DKO than parental cells. These data are consistent with the results of our DNA analysis, which showed evidence of CpG methylation in HCT116 parental cells but was largely lacking in DKO cells. Elevated IκBα expression in DKO cells thus may reflect differences in both DNA methylation as well as altered chromatin configuration at the IκBα locus.
The methylation of CpG dinucleotides and the posttranslational modification of histones represent two important epigenetic mechanisms for the regulation of gene expression levels. Epigenetic regulation of gene expression has received great attention in developmental biology where it controls essential processes such as genetic imprinting and X chromosome inactivation. Epigenetic factors have also been extensively investigated in cancer biology, where the methylation of CpG dinucleotides in promoter regions of tumor suppressor genes has been implicated in the acquisition of the cancer cell phenotype. In this work, we have addressed a hitherto little-studied area in epigenetics: regulation of the NF-κB system by CpG methylation and histone modifications at the IκBα promoter. Since NF-κB is a dominant controller of innate immunity, these findings represent a novel recognition of epigenetic regulatory mechanisms of immune homeostasis in the intestinal epithelium.
Recent studies have revealed some new insights into how epigenetic factors such as DNA methylation can affect immunity. For example, in T lymphocytes, the transcription factor FOXP3 results in the acquisition of a regulatory phenotype, and it has been shown that methylation of a CpG island in its promoter can suppress expression of FOXP3, with resultant loss of this regulatory phenotype (12, 25). Other work has shown that the expression of the cytokine IL-10 is regulated by DNA methylation of the IL10 locus (17, 33). In a model of T lymphocyte differentiation into T helper types 1 or 2, it has been shown that high levels of secretion of interferon-γ, which is characteristic of the T helper-1 phenotype, is accompanied by loss of CpG methylation of the IFNG locus, favoring its expression (22). Our work has shown that the immune functions of intestinal epithelial cells can also be affected by CpG methylation, albeit by an indirect mechanism.
In HCT116 cells, we showed that the expression of IL-8 was in fact lower in cells lacking DNMT, as a result of higher expression of IκBα, the endogenous inhibitor of NF-κB, the transcription factor that controls its expression. Lower induced IL-8 expression in HCT116 DKO compared with parental cells was accompanied by a modest reduction in the ability of media conditioned by these cells to stimulate monocyte chemotaxis. It is likely that monocyte chemotaxis, an important physiological process for the inflammatory response and also important for the maintenance of pathological inflammation, is regulated by multiple chemokines in addition to IL-8 (38). Moreover, monocytic cells infiltrate intestinal carcinomas, and significant evidence points toward their potential to promote tumorigenesis (31). Therefore, our data suggest that the inhibition of DNA methylation should be explored as a therapeutic approach to inhibit the chemotaxis of monocytic and possibly other immune cells whose recruitment depends on the secretion of NF-κB-regulated chemokines from intestinal epithelial cells.
HCT116 DKO cells expressed less cFLIP than did parental cells. cFLIP is another NF-κB-regulated protein that is a dominant inhibitor of death receptor-mediated apoptosis. Assessed by an active caspase 3 assay, HCT116 DKO cells responded to TNF-α stimulation with increased apoptotic signaling compared with parental cells, an observation that is consistent with lower cFLIP levels. The sensitivity of intestinal epithelial cells to undergo apoptosis is an important factor influencing their homeostatic role (2). Thus, if intestinal epithelial cells die more readily, barrier function is compromised. Conversely, if intestinal epithelial cells are resistant to death, neoplasia may be more likely to develop (15). Prior work using DNA methylation inhibitors such as 5-azadeoxycytidine has shown that those drugs can induce cancer cell apoptosis (41). It is likely that many mechanisms mediate the proapoptotic effect of those drugs, but our findings suggest that one relevant mechanism may be the increased expression of IκBα and its inhibitory effect on NF-κB, an important inhibitor of intestinal epithelial cell apoptosis.
Bisulfite sequencing of DNA cloned from the HCT116 cell line showed that some cells exhibited little or no methylation of the IκBα CpG island, whereas others had intermediate levels of methylation and others had almost complete methylation of this locus. The explanation for this heterogenous pattern is unclear, but prior reports of the methylation patterns of other genetic loci have also revealed significant heterogeneity (5). HCT116 DKO cells exhibited very few methylated CpG dinucleotides in the IκBα promoter island and no heavily methylated clones. This finding is consistent with the near absence of DNA methylation in this clone or could possibly indicate that this clone was derived from a parental cell with an unmethylated IκBα CpG island. Analysis of histone-DNA interactions using native ChIP revealed that the histone marks Ac-H3K14, H3K4(Me2), and H3K9(Me3) bound in higher abundance to the unmethylated DNA of HCT116 DKO compared with parental cells. The increase in association of acetylated histones occurred in both the coding region (Exon 1) as well as the promoter of IκB-α, suggesting that histone acetylation occurs throughout the entire gene. These differences were accompanied by significantly higher expression of IκBα. The magnitude of IκBα overexpression in HCT116 DKO relative to parental cells was approximately twofold at the mRNA level, with a comparable or slightly greater difference in protein levels. However, TNF-α-induced NF-κB activation was almost completely abolished in HCT116 DKO cells. Although IκBα levels did fall following TNF-α stimulation in HCT116 DKO cells, it appears not to have fallen to a sufficiently low level to liberate NF-κB for nuclear translocation. The strong effect of DNA methylation on TNF-α-induced NF-κB activity is mirrored in the greatly reduced induction of IL-8 and cFLIP in HCT116 DKO cells.
One important limitation to our work is the fact that the cell lines chosen for our studies were derived from a colon cancer, and therefore the extent to which our findings can be extrapolated to nonneoplastic intestinal epithelium is as yet uncertain. It is well understood that CpG island methylation is increased in many colon cancers compared with the normal colon and can also increase with age (18). Recent studies have discovered that chronic inflammation can also be associated with increased CpG island methylation (19). Significant evidence now supports the role of epigenetic factors in intestinal tumorigenesis, and there is increasing understanding of the role and complexity of epigenetics in inflammation and inflammation-associated carcinogenesis (26). Therefore, the finding of epigenetic regulation of the IκBα-NF-κB system in intestinal epithelial cells provides additional important mechanistic insight into these processes that requires further investigation.
This work was supported by grants from Science Foundation Ireland.
The authors report no conflicts of interest.
- Copyright © 2010 the American Physiological Society