INTRODUCTION

Colon cancer is one of the most commonly occurring tumors and a major cause of cancer-related deaths world-wide. Colon cancer cell lines Colo-3201, 2, 3 and SW11164, 5, 6 are frequently used in molecular biological experiments. The transcription of CDKN2A7, p21WAF1 8 and adenomatous polyposis coli (APC)9 genes are down-regulated and c-myc proto-oncogene is over-expressed10 in the Colo-320 cell line.

CDKN2A inhibits the catalytic activity of the cyclin-dependent kinases 4 (CDK4) /cyclin D complex and also blocks the G1/S transition in cells11, 12. p21WAF1 is one of the CIP/KIP family, which inhibits late G1/S check point kinases. The increased expression of p21WAF1 may induce growth arrest in transformed cells13, 14, 15. c-myc is a transcription factor whose normal function is to promote cell proliferation8, 16.

Epigenetic modifications, mainly DNA methylation17 and changes of histones, are now recognized as additional mechanisms contributing to malignant phenotypes18. Several studies have demonstrated that the inactivation of CDKN2A in human colon tissue may be due to de novo methylation of its 5' promoter-associated CpG island19, 20, 21, 22, 23.

Hyperacetylation of histones has been shown to open chromatin and is required for transcriptional activation24. Because the inhibitory effects of complex endogenous genes may play a significant role in the G1-S progression of the cell cycle, histone deacetylases (HDAC) inhibitors, trichostatin A (TSA, a hybrid polar compound specific inhibitor )24, 25, 26 and sodium butyrate (butyrate, a short chain fatty acid, SCFA)27, 28, 29 have been considered as candidates for the treatment of cancer20.

Previous studies focused on aberrant methylation or acetylation alone in a single gene. However, little is known about the effect of the regulation of methylation and histone acetylation on the expression of several tumor-associated genes including the tumor suppressor gene and proto-oncogene in human colon cancer. The DNA methylation status, gene-associated-histone acetylation pattern and the transcriptional level of tumor-associated genes in the presence of combined 5-aza-dC and HDAC inhibitors in the Colo-320 and SW1116 cell lines remain unclear. Furthermore, we want to know whether 5-aza-dC or TSA induces over-expression of proto-oncogene while regulating the transcription of tumor suppressor genes. The aim of this study is to gain a better understanding of the effect of the regulation of DNA methylation and histone acetylation on the regulation of tumor suppressor gene and proto-oncogene transcription.

MATERIALS AND METHODS

Cell culture

The colon cancer-derived cell lines Colo-320 and SW1116 were maintained by serial passages in RPMI1640 containing 10% heat-inactivated FCS, 100 U/ml penicillin and 100 μg/ml streptomycin, and incubated at 37°C, 5% CO2-95% air using standard tissue culture incubators as described previously30.

Treatment with 5-aza-dC, TSA and butyrate

5-aza-dC is an inhibitor of DNA methyltransferase (Dnmt 1) 31, but TSA and butyrate are HDAC inhibitors. To assess the restoration of CDKN2A, p21WAF1, APC, p53, p73, c-myc, c-Ki-ras and survivin genes expression by 5-aza-dC, TSA or butyrate treatment, the colon cancer cell lines were exposed to different concentrations (2 μM, 5 μM or 10 μM) of 5-aza-dC (Sigma, St. Louis,MO) alone for 24 h and 72 h; 1 μM TSA or 5 μM butyrate (Sigma, St. Louis, MO) alone for 24 h. For the combined treatment of drugs, cells were incubated with 10 μM of 5-aza-dC for 2 d and then with 1 μM of TSA or 5 μM of sodium butyrate for 24 h. The control cultures were treated with phosphate-buffered saline (PBS) or ethanol (for TSA treatment, because TSA can only be dissolved in PBS containing ethanol). DNA and RNA were extracted at various time points.

Cell viability assays

The 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenytetrazolium (MTT) assay measures both cytostatic and cytotoxic responses to drug treatment and thus provides an accurate overall measure of drug sensitivity. Cell viability assays were performed in quadruplicate wells of 24-well plates 3 d after the treatment of 5-aza-dC or TSA or sodium butyrate. MTT (Calbiochem, San Diego, CA) assay was used to assess drug-induced cell growth inhibition and cytotoxicity. No significant cytotoxic damage was observed during the drug treatment.

Bisulfite modification and methylation-specific PCR (MSP) for CDKN2A promoter

To address whether DNA methylation of CDKN2A changed during the drug treatment, we carried out bisulfite modification32 and MSP to define a CpG-rich region in the CDKN2A gene promoter.

Bisulfite converts unmethylated cytosine residues to uracil, but methylated cytosines remain nonreactive. PCR amplifies uracil as thymine while methylated cytosines can only be amplified as cytosines. Genomic DNA treated by bisulfite was amplified with CDKN2A promoter gene fragment specific primers. PCR reaction buffer contained 0.1 mM dNTP, 2.0 mM MgCl2, and 0.5 μM primers. PCR product was directly loaded on to 3% agarose gels and electrophoresed. The gel was stained with ethidium bromide and directly visualized under UV illumination. The sequences and PCR program are shown in Tab 1. Furthermore, wild-type CDKN2A primers were used to monitor the complete conversion of DNA obtained in the bisulfite reaction.

Table 1 Sequences and PCR programs of primers for RT-PCR, sequencing and ChIP-PCR

Chromatin immunoprecipitation (ChIP) assay

A ChIP assay kit from Upstate Biotechnology was used according to the manufacturer's protocol and Richon's report33. Colo-320 and SW1116 cells, mock-treated and treated with drugs, were plated at a density of 10×106 cells/T25 flask. p21WAF1- and CDKN2A -specific primers were used to carry out PCR. The density of bands in ChIP-PCR was quantitated using a Molecular Dynamics PhosphorImager (Nucleo Tech Inc, San Mateo, CA). The sequences of one set of primers for CDKN2A and two sets of primers for p21WAF1 PCR, as well as the PCR condition are shown in Tab 1. The 1st set of primers was used to amplify -576 to -293 and the 2nd set of primers was used to amplify -51 to +77 of p21WAF1 promoter and exon 1, which contained the transcription factor E2A binding sites. The 395 bp band corresponds to the promoter of CDKN2A (21 CpG sites, positions -494 to -101)34.

Reverse transcription polymerase chain reaction (RT-PCR)

Total RNA was extracted by using a commercial kit (Trizol) according to the manufacturer's instructions (Gibco BRL). RT reactions using 5 μg of total RNA were performed with Superscript II reverse transcriptase (Life Technologies, Inc.). The mRNA transcription levels of p53, p73, c-Ki-ras and survivin genes were determined by RT-PCR. The sequences and PCR reaction for each primer are shown in Tab 1. For the control of RT-PCR, a 612 bp fragment of β-actin cDNA was also amplified.

Real-time quantitative PCR

The mRNA levels of CDKN2A, p21WAF1, APC and c-myc were measured using a real-time quantitative PCR system. Relative quantitation using the comparative Ct method with data from the ABI PRISM 7700 Sequence Detection System (version 1.6 software) was performed according to the manufacturer's protocol. The primers and Taqman fluorogenic probes for CDKN2A, p21WAF1, APC and c-myc were provided by Shenyou Company, Shanghai. The sequences of forward and reverse primers and the probes are shown in Tab 2. Real-time PCR was also performed with primers and Taqman β-actin probes to normalize each of the extracts for amplifiable human DNA. The results were expressed as the ratio of copies of CDKN2A, p21WAF1, APC and c-myc to β-actin, respectively. The Ct values were measured, and the average Ct of triplicate samples was calculated. Alteration of mRNA expression was defined as a 3-fold difference in the expression level after treatment, relative to that before treatment 35.

Table 2 Sequence of primers and probes for real-time PCR

Flow cytometric (FCM) detection of cell cycle

Cell cycle analysis was carried out by FCM36. In brief, a total of 1×108 cells were removed from the treated and mock-treated cultures at specified time points. The cells were washed twice with PBS and fixed in ice-cold ethanol for 1 h. The samples were concentrated by removing ethanol and treated with 1% (v/v) Triton X-100 (Sigma, USA) and 0.01% RNase (mg/ml, sigma, St. Louis, Mo.) for 10 min at 37°C. The staining of cellular DNA was performed with 0.05% propidium iodide for 20 min at 4°C in darkness. The cell cycle distribution was detected using a flow cytometer (Model FACSCALIBAR, BD, USA) and 10,000 cells were analyzed with the MultiCycle software package (Phoenix, San Diego, CA, USA).

Statistical analysis

Differences in the cell cycle between those cells treated with drugs and those mock-treated were analyzed by the Fisher exact test.

RESULTS

Demethylation of promoter and the restoration of expression of CDKN2A gene by 5-aza-dC

In the first part of this study, we examined the possibility of methylation in regulating the expression of CDKN2A in Colo-320 and SW1116 cells. As shown in Fig 1, mock-treated Colo-320 and SW1116 cells showed a positive 150 bp band and 151 bp band for methylated and unmethylated specific primer sets for CDKN2A respectively, indicating that the CDKN2A gene is partially methylated in the tested cell lines. The methylated bands of CDKN2A gene in the cells treated with 10 μM of 5-aza-dC for 24 h were consis-tently lower than those of the mock-treated cells (Fig 1). Thus, unmethylated product was significantly higher in the 5-aza-dC treated cells. The data from this study showed that untreated Colo-320 and SW1116 cells were methylated at some CpG sites in most alleles. After treatment with 10 μM of 5-aza-dC, the cytosine (C) in CpG changed to thymine (T) in some sites.

Figure 1
figure 1

Methylation analysis of CDKN2A gene promoter by MSP. MSP was performed with the specific primers described in the Materials and Methods.

As shown in Tab 3, 5-aza-dC in a higher concentration (10 μM) induced transcription of CDKN2A at 24 h, but not in a lower concentration (2 μM or 5 μM). In contrast, at 72 h post-treatment, the level of CDKN2A mRNA of cells exposed to 5-aza-dC was less than that at 24 h.

Table 3 Alteration of mRNA expression of CDKN2A, p21WAF1, APC, c-myc and β-actin in Colo-320 and SW1116 cells, using real-time quantitative PCR

However, the transcription of the CDKN2A gene can be induced by 5-aza-dC alone but not by TSA alone or the combination of 5-aza-dC and TSA. Also the treatment of TSA or sodium butyrate failed to induce the acetylation of histones H3 and H4 in colon cancer cells (Fig 2C). Thus, the above results indicate that DNA methylation but not histone acetylation is the major regulating mechanism for CDKN2A expression in human colon cancer cell lines.

Figure 2
figure 2figure 2figure 2figure 2

TSA and sodium butyrate (NaBu) but not 5-aza-dC induced accumulation of acetylated histones H3 and H4 in p21WAF1 gene but not CDKN2A. Soluble chromatin from Colo-320 cells treated with or without TSA or NaBu or 5-aza-dC was immunopreciptated with antiacetylated histones H3 and H4 antibodies. The 1st (A) and the 2nd (B) sets PCR primers for the regions of p21WAF1, and (C) were PCR primers for the promoter of CDKN2A as indicated in Materials and Methods. (D) The Figs in A and B were scanned and quantified by Molecular Dynamics PhosphorImager. The ratio between input DNA and precipitated DNA was calculated for each treatment and primer set. The folds increased after treatment with TSA or NaBu or 5-aza-dC were calculated from the indicated ratios.

5-aza-dC but not TSA or sodium butyrate increased the transcription of the APC gene

To identify whether the transcription level of APC is regulated by DNA methylation or histone acetylation or both in human colon cancer cells, APC mRNA level was assayed. Colo-320 and SW1116 cells were cultured with or without 5-aza-dC or TSA or sodium butyrate for 24 h and 72 h. An incubation of 24 h with 5-aza-dC resulted in the accumulation of APC mRNA, which remained at a higher level during a 72 h incubation period. The APC mRNA levels were normalized with β-actin mRNA (Tab 3). The effect of 5-aza-dC on the expression of APC was high even at a very low dosage (2 μM for Colo-320 and 5 μM for SW1116 cells), suggesting that methylation-induced silencing of this gene was the primary event. The restoration of APC expression by 5-aza-dC confirmed a causal relationship between DNA hypermethylation and APC silencing in colon cancer cell lines Colo-320 and SW1116.

No obvious change was seen during the treatment with TSA or sodium butyrate. The expression of the housekeeping gene β-actin was used as a control to ensure the similarity in quality and quantity of RNA samples amplified by real-time RT-PCR.

Either TSA or sodium butyrate but not 5-aza-dC induced the acetylation of p21WAF1 gene-associated histones and the re-expression of p21WAF1

To determine whether histone acetylation affects p21WAF1 transcription, and whether a functional interaction occurs between p21WAF1 and TSA or sodium butyrate treatment, ChIP-PCR was performed. The input was 10% of the amount of chromatin which was used for immunopre-cipitation.

As shown in Fig 2, the densities of bands of p21WAF1 gene-associated acetylated histone H4 and H3 were higher in chromatin extracted from either TSA or sodium butyrate treated Colo-320 cells than that from mock-treated cells. 285- and 128 bp fragments of p21WAF1 promoter region and exon 1 were amplified (Fig 2A and 2B). In the PCR product containing transcription factor E2A binding sites of promoter region, the positive bands that were not detected in controls were found after being cultured with TSA or sodium butyrate.

On the other hand, sodium butyrate or TSA or both combined with 5-aza-dC greatly increased the p21WAF1 transcription level in the cells. As shown in Tab 3, after treatment with 5-aza-dC alone, there was no significant difference between the expression of this gene in cells treated with and without drugs. In addition, in Colo-320 cells, p21WAF1 expression increased to a higher level than that in SW1116 cells.

Taken together, the above results indicated that TSA or sodium butyrate activated the transcription of p21WAF1 through acetylating histones H4 and H3 associated with the p21WAF1 promoter.

5-aza-dC or TSA or sodium butyrate treatment failed to induce expression of p53, p73, c-myc, c-Ki-ras and survivin in Colo-320 and SW1116 cells

To further define the modification status of proto-oncogene c-myc overexpression in colon cancer cells, we attempted to clarify whether c-myc expression is altered after treatment with Dnmt inhibitor or HDAC inhibitors. Although the significant overexpression of c-myc and c-Ki-ras has been found in mock-treated Colo-320 and SW1116 cells, our current study revealed that almost no change in transcription was seen when these two cell lines were treated by 5-aza-dC or TSA or sodium butyrate. In other words, no evidence was found that 5-aza-dC and TSA or sodium butyrate could induc the accumulation of c-myc gene mRNA (Tab 3). Taken together, the 5-aza-dC and TSA or sodium butyrate-induced changes in methylation and histone acetylation were localized at specific regions of the chromatin. Other mechanisms might explain a possibility in controlling the activity of the c-myc gene in colon cancer cell lines Colo-320 and SW1116. In addition, the effect of regulation of epigenetic modification on the expression of p53, p73, c-Ki-ras and survivin was not found (data not shown).

TSA or sodium butyrate induces cell cycle arrest

To address whether demethylation or acetylation of p21WAF1 affects the cell cycle, FCM was performed as described above. As expected, the data demonstrated a significant accumulation of cells in the G1 and G2 phases, and a decrease in the S-phase, thus indicating that either TSA or sodium butyrate induced cell cycle arrest in these two cell lines. Tab 4 further elucidated the association between hypoacetylation and the cell cycle, showing that the ratio of the cell number in G1/G0 increased significantly, and that in the G2/M and S-phases it decreased. Interes-tingly, 5-aza-dC-treated SW1116 cells did not present comparable cell cycle arrest although 5-aza-dC induced the restoration of CDKN2A expression; also there was no correlation between CDKN2A gene expression and changes in the cell cycle of SW1116 cells (see representative examples in Fig 3).

Table 4 TSA and sodium butyrate induce the arrest of G1 to S phase
Figure 3
figure 3figure 3figure 3figure 3figure 3figure 3figure 3

FCM analysis of SW1116 cells. TSA or sodium butyrate or combination treatment of TSA/sodium butyrate with 5-aza-dC induces the arrest of G1 to S phase in SW1116 cells. (A) mock treated, control; (B) ethanol control; (C) 5-aza-dC treated; (D) TSA treated; (E) sodium butyrate treated; (F) TSA + 5-aza-dC treated; (G) sodium butyrate + 5-aza-dC treated.

Taken together, these results suggested that due to the lack of p21WAF1 expression in human colon cancer cells, and hence an absence of G1-phase arrest in the cells, the genes were re-expressed after treatment with TSA or sodium butyrate. Similar results were obtained from another colon cancer cell line, Colo-320 (data not shown).

DISCUSSION

DNA methylation and chromatin modification are two global mechanisms for regulating gene expression. CpG island chromatin is found to contain highly acetylated histones H3 and H437. The process of DNA methylation mediated by Dnmt1 may depend on the generation of an altered chromatin status via histone deacetylase activity 38. Many investigators showed that genes silenced by the promoter hypermethylation could be reactivated by treatment with the demethylating drug 5-aza-dC, either alone or in combination with an HDAC inhibitor31, 39, suggesting that DNA methylation and chromatin architecture act together to silence genes. Furthermore, TSA induces G0→G1 cell cycle arrest40.

Several studies indicate that CDKN2A is hypermethylated in colon cancer cells41, and 5-aza-dC-induced growth inhibition may result from the release of methylation silencing of the cell cycle regulatory genes, CDKN2A42. However, it remains unclear whether histone acetylation regulates the transcription of CDKN2A and APC genes. Our finding that 5-aza-dC but not TSA or sodium butyrate induces the transcription of CDKN2A and APC, suggests that DNA methylation rather than histone acetylation is the major regulation mechanism for these two genes in Colo-320 and SW1116 cells. We noted that 5-aza-dC enhanced CDKN2A transcription, but the combination of 5-aza-dC and TSA did not induce more expression of CDKN2A. It is possible that apoptotic cell deaths occurred when cells were treated by a combination of demethylation and TSA 43. However, for p21WAF1, either TSA or the combination of TSA and 5-aza-dC exhibits a 17- to 30-fold (Colo-320 cells) or 3- to 6-fold (SW1116 cells) increase in trans-cription, suggesting that synergic gene activation occurs (Tab 3). Regarding the mechanism of sensitization mentioned above, these data indicated that the deme thylation of the gene promoters and the activity of 5-aza-dC could have complex effects on the localization and activity of HDACs that had not been previously contemplated44.

Butyrate is known to alter the expression of a variety of genes by modulating histone acetylation45, 46, 47. Both sodium butyrate and TSA were shown to inhibit HDAC, leading to hyperacetylation of selective histone proteins such as histone H446. However, most previous studies have demonstrated that the transcription of p21WAF1 could be strongly enhanced by TSA in the colon cancer cell line, HT2948, but have not elucidated the expression status of this gene during the treatment of 5-aza-dC or HDAC inhibitors in the human colon cancer cell lines Colo-320 and SW1116. As summarized in Tab 3 and 4, the hypera-cetylation of histone but not demethylation induced the re-expression of p21WAF1 and a G1-phase arrest of Colo-320 and SW1116 cells. We also implied that either TSA or sodium butyrate or in combination with 5-aza-dC, but not 5-aza-dC alone, could induce p21WAF1 expression. We used real-time quantitative PCR based on TaqMan methodology to analyze quantitatively the expression of these genes in routinely processed cells and demonstrated that the measurements of gene expression could be reliably and accurately conducted in such cells.

The inhibition of proliferation demonstrated a significant accumulation of cells in the G1 and G2 phases, and a decrease of cells in the S-phase of the two TSA-treated cell lines. TSA and sodium butyrate-induced growth inhibition could result from the release of hypoacetylation silencing of the cell cycle regulatory genes such as p21WAF1. We also examined whether the re-expression of CDKN2A induced by promoter hypomethylation could induce the arrest of the cell cycle similar to that by p21WAF1 in human colon cancer cells. The result was negative. Data from this report supports the idea that p21WAF1 expression is more important than CDKN2A for the regulation of the cell cycle in the human colon cancer cell lines Colo-320 and SW1116.

Although reduced levels of methylation of genes including c-myc in human tumors have also been reported, it has not been shown convincingly that they are indeed responsible for increased levels of gene expression rather than merely a secondary characteristic observed in cancer cells. Over-expression and abnormal intracellular location of the product of the proto-oncogene c-myc in colon dysplasia and neoplasia may be related to alterations in epigenetic mechanisms which controll the function of this gene49. p73 is a new member of the p53 family. The transcriptional silencing of the p73 gene by hypermethylation of a CpG island was observed in several leukemia and lymphomas 50. Survivin is the first apoptosis inhibitor described to-date to be expressed in G2-M, in a cell cycle-dependent manner45 and is detected in all cases of normal colon mucosa. Up to now, little is known about the relevance between the methylation and transcriptional level of c-myc, p5347, p73 and survivin genes during colon carcinogenesis or human colon cancer cell lines. Therefore, the coexistence of methylated and unmethylated status of several tumor suppressor genes and proto-oncogenes in the same colon cancer cell line, reflects the in vitro situation and has some functional significance.

The effect of the promoter methylation of several tumor- associated genes and histone acetylation of p21WAF1 and CDKN2A transcription have been analyzed in human colon cancer cell lines Colo-320 and SW1116. In this study, the results from RT-PCR and real-time RT-PCR indicate first that CDKN2A, APC, p53, p73, survivin and c-myc, c-Ki-ras are expressed in Colo-320 and SW1116 cell lines. Furthermore, epigenetic modification did not seem to be involved in the transcription regulation of p53, p73, survivin and c-myc, c-Ki-ras genes in these two cell lines.

One of the most important findings from this study is that neither DNA methyltransferase inhibitor nor HDAC inhibitor regulates the transcription of c-myc and c-Ki-ras proto-oncogenes in colon cancer cell lines Colo-320 and SW1116, although c-myc51 and c-Ki-ras52 were over-expressed in most human colon cancer. Previous clinical trials indicated that 5-aza-dC was devoid of anti-tumour activity in adult patients with colon cancer53, 54, 55. The data in present study suggests that the reason for its failing in colon cancer treatment is not c-myc over-expression from demethylation.

In conclusion, the major findings of this present study support the concept that histone acetylation but not methylation is the major mechanism in the regulation of p21WAF1 gene expression, whereas methylation regulates the expression of CDKN2A and APC. The transcription level of p53, p73, c-myc, c-Ki-ras and survivin tumor associated genes are regulated by either methylation or histone acetylation in human colon cancer cell lines Colo-320 and SW1116. Our results not only enhance our understanding of the molecular epigenetic mechanisms underscoring colon carcinogenesis, but also facilitate the development of diagnostic tools based on the DNA methylation profile and histone acetylation status for the early diagnosis of human colon cancer.