With the informed consent of all patients and approval of the ethics committee, the samples of tumors were collected from HCC patients (n = 30) during operation. The pathological classification of tumor tissues was carried out, and the stage of each case was graded according to the WHO classification[15]. Total genomic DNA was extracted from frozen tissue specimens (50-100 mg) according to a standard protocol with some modifications[16,17]. Frozen pulverized powders of the specimens were re-suspended with 2 mL lysis buffer containing 50 mmol/L Tris-HCl pH 8.0, 50 mmol/L EDTA, 10 g/L SDS, 10 mmol/L NaCl plus 100 μg/mL boiling-treated RNase A (Sigma, USA). Following 1-h incubation at 37 °C, proteinase K (Roche, USA) was added to the cellular lysates for a final concentration of 100 μg/mL, and the digestion was carried out at 55 °C for 2 h. Organic extractions with a half volume of phenol/chloroform/isoamyl alcohol (1:1:0.04) were repeatedly carried out until no visible interphase remained after centrifugation. DNA was precipitated from the aqueous phase in the presence of 0.3 mol/L NaOAc (pH 7.0) and two and a half volumes of ethanol, washed once in 700 mL/L ethanol, dissolved at 65 °C for 30 min with 0.2-0.4 mL TE (10 mmol/L Tris-HCl pH 7.4 and 1 mmol/L EDTA) and stored at 4 °C until use. The DNA concentrations were calculated according to the A_{260 nm} readings.

We used “drug resistance” as the key word to search from the NCBI LocusLink database (http://www.ncbi.nlm.nih.gov/ LocusLink/list.cgi). The sequence of the coding region (63 targets) plus 5 kb segments at each ends were downloaded, and subjected to search for the existence of the CpG island. The parameters were set as OBS/EXP (the minimum average observed to expected ratio of C plus G to CpG) = 0.6, MINPC (the minimum average percentage of G plus C) = 50, and LENGTH (the minimum length that a CpG island has to be) = 200 bp (http://www.ebi.ac.uk/emboss/cpgplot/). Thirty six genes were found containing the promoter CpG island (Table 1). Among this list 14 genes were selected for the methylation profiling, as they represent in different categories for the functions and properties (Table 2). The primer pairs (Table 3) for methylation specific polymerase chain reaction (MSP) in this report were designed according to the same principle with assistance of the web server for identification of the CpG islands (http://http://www.ebi.ac.uk/emboss/cpgplot/index.html) and the primer design software (http://micro-gen.ouhsc.edu/cgi-bin/ primer3_http://www.cgi) (Table 3).

The methylation status of the promoter CpG islands of the 14 genes in all sample DNAs were analyzed by MSP on the sodium-bisulfite converted DNA[16,17]. The primer pairs for MSP and the genes’ genome access on GenBank are detailed in Table 3. In detail, 2 μg DNA in 50 μL TE was incubated with 5.5 μL of 3 mol/L NaOH at 37 °C for 10 min, followed by a 16 h treatment at 50 °C after adding 30 μL of fresh-prepared 10 mmol/L hydroquinone and 520 μL of freshly prepared 3.6 mol/L sodium-bisulfite at pH 5.0. The DNA was desalted using a home-made dialysis system with 10 g/L agarose. Then, the DNA (approximately 100 μL) was denatured at 37 °C for 15 min with 5.5 μL of 3 mol/L NaOH, followed by ethanol precipitation with 33 μL 10 mol/L NH₄OAc and 300 μL ethanol. After washing with 700 mL/L ethanol, the gently dried DNA pellet was dissolved with 30 μL TE at 65 °C for 10 min. The DNA sample was finally stored at - 20 °C until further use. PCR reaction was carried out in a volume of 15 μL with 50 ng or less template DNA with JumpStart Taq polymerase (Sigma, USA). An initial denaturation step at 94 °C for 4 min was followed by 35 cycles of reaction at 94 °C for 20 s, variable annealing temperature (Table 3) for 20 s and 72 °C for 25 s, and a final extension at 72 °C for 5 min. The PCR products were separated by 12 g/L ethidium bromide containing agarose gel electrophoresis with 1 × Tris Acetate EDTA (TAE) buffer and visualized under UV illumination. To verify the PCR results, representative bands from each target were gel-purified and cloned into T-vector (Buocai, Shanghai, China), followed by automatic DNA sequencing provided by Buocai. Only verified results are presented in this report.

The M. Sss I treated DNA from the normal liver tissue was used in the MSP procedure as the control template. The DNA from the liver tissue of the healthy liver donor[14,16] was batch cleaved with EcoRI, followed by M. Sss I treatment overnight according to the manufacturer’s instruction (New England Biol., Boston, USA). The purified DNA was bisulphate treated as usual, and subjected to MSP with the primer pairs for 11 genes where no methylated alleles were detected in both cell lines and HCC tissues.

The established cell lines, BEL-7402 (human hepatocellular carcinoma cell line, No. TCHu68, Cell Bank in Shanghai), SMMC-7721 (human hepatocellular carcinoma cell line, No. TCHu13, Cell Bank in Shanghai), Hep3B (human hepatocellular carcinoma cell line, ATCC Number: HB-8064), HepG2 (human hepatocellular carcinoma cell line, ATCC Number: HB-8065), HCCLM3 (human hepatocellular carcinoma cell line, Cell Bank in Zhongshan Hospital, Shanghai), and L02 (immortalized normal liver cell line, No GNHu6, Cell Bank in Shanghai) were cultured in DEME medium plus 100 mL/L new born calf serum at 37 °C in a 50 mL/L CO₂ atmosphere.

Total RNA from various HCC cell lines was extracted using the TriPure isolation reagent kit (Roche, USA). Reverse transcription was subsequently carried out using the superscript II RNase H-reverse transcriptase kit. β -actin was used as an internal control in separate reaction. The GSTpi cDNA primers (sense: 5’-GGAGACCTCACCCTGTACCA-3’; anti-sense: 5’-GGGCAGTGCCTTCACATAGT-3’) were the same as in a previously published paper[19]. The primer sequences and reaction conditions are listed in Figure 1. RT-PCR products were visualized under UV illumination after electrophoresis in 1.2% agarose, followed by the densitometric quantification.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 The hypermethylation for each gene in each of HCC patient samples

The methylation data were dichotomized as 1 for the co-existence of the methylated and unmethylated alleles; 2, for methylated allele only; and 0 for the unmethylated for both alleles to facilitate statistical analysis using contingency tables. The statistic analyses for the association between the methylation profile of the gene and each of the clinical-pathological parameters were carried out with the statistics package (http://www.R-project.org/). Both Pearson’s Chi-square test with Upton’s adjustment and Fisher’s exact test (http://www.R-project.org/) were used to deal with the sample cells with the low expected values. The relative frequency with a 95% confidence interval (P < 0.05) for a binomial distribution was calculated for the tumors and the paired non-cancerous tissues.

Sixty three genes were revealed from NCBI LocusLink database from the searching with “drug-resistance” as the key words (http://www.ncbi.nlm.nih.gov/LocusLink/list.cgi). To identify the genes falling into the category of CpG island containing genes, the sequence of the coding region plus 5 kb each upstream and downstream was downloaded and subjected to the analysis with CpG island identification software (http://http://www.ebi.ac.uk/emboss/cpgplot/). Forty four genes were found containing one or more CpG islands, but only 36 genes really had the promoter embedded in a CpG island (Table 1). Hence, transcription of these 36 genes is likely to be subjected to the control of the DNA methylation status. From these genes, 14 genes representing different functions and mechanisms in “drug-resistance” (Table 2) were selected for methylation profiling; three genes encoding the various membrane transporters: ABCG2, CFTR, and RALBP1, two encoding the transcription factors, ATF2, and TP53, four encoding the enzymes mediating the drug activation or inactivation: CAT, GSTpi, TOP2B and DCK, two encoding the proteins involved in signaling or cell-cycle progression: RAF1 and SKP2, and three with unknown underlying mechanisms: B2M, OCLN and SPF45.

MSP is the most popular method for methylation profiling of any given targets in human cancers for both its easiness and sensitiveness. Based upon the distinguished difference in sensitivity of the methyl-cytosine from the cytosine to be converted to uridine by bisulphate treatment: the later is amenable, but the former is resistant, the methylation status of the target fragments can be determined by parallel PCR reactions with one pair of the primers in which the CG in the wide-type sequence is maintained for detection of methylated allele, and the other in which the CG is replaced with TG for the unmethylated allele. However, inherent with the PCR-type assays, there are compelling reasons to eliminate both false positive and negative PCR reactions. As 13 genes in this list except for the GSTpi gene had not been subjected to MSP previously, we carried out optimization of the PCR reaction with a panel of five established cell lines. For those targets, only the unmethylated allele was detected, MSP with the DNA from the normal liver tissue was carried out, showing that all 14 genes were unmethylated. To ensure that failure to detect the methylated allele was not false negative artifact in PCR reaction, we in vitro methylated the DNA from the normal liver with M. Sss I methyl-transferase that was capable of transferring methyl to the fifth carbon atom of the cytosine in the CpG dinucleotides, followed by methylation profiling. As shown in Figure 2, while no methylated allele was detected with the methylation targeted primers with the liver DNA, the M. Sss I treated DNA led to the positive PCR reaction, indicating that the PCR condition used for the tumor samples was capable of detecting the methylated targets, including ABCG2, ATF2, B2M, DCK, RAF1, RALBP1, SPF45, SKP2, TP53, and TOP2B genes (panels 1-10, respectively). As the rest four genes were heterozygously methylated in some of liver cancer cell lines, it indicated that the PCR reactions specific to the methylated allele was acceptable (for instance, lane 11 for OCLN, Figure 2), the in vitro methylation by M. Sss I testing was exempted. To eliminate the false negative results, the representative PCR bands were subjected to T-cloning and sequencing verification. Only the data, the bands of which had the correct identity in sequence, were regarded informative.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 MSP testing with the M. Sss I treated normal liver tissue DNA. The in vitro methylated DNA from the normal liver tissue was used for MSP along with the untreated. Panels: 1, ABCG2; 2, ATF2; 3, B2M; 4, DCK; 5, RAF1; 6, RALBP1; 7, SKP2; 8, SPF45; 9, TOP2b; and 10, TP53. The PCR conditions for OCLN gene were assessed by PCR in three liver cancer cell lines, 7721, 7402, and Hep3B (panel 11). U, unmethylated allele; M, methylated allele, NL, normal liver tissue; *the DNA size marker: DL-2000.

By MSP analysis, 13 genes displayed a uniform unmethylated status in the liver tissues from all four healthy donors, while the RAF1 gene was heterozygous in one case, but fully unmethylated in the rest three cases. In all 30 cases of HCC, the following 11 genes: ABCG2, ATF2, B2M, DCK, RAF1, RALBP1, SPF45, SKP2, TOP2B, OCLN and TP53 maintained the same unmethylated status in both HCC tissues and the paired non-cancerous tissues (Figure 3), indicating no DNA methylation mediated changes in the control of the expression of these 11 genes in HCC. The remaining three genes displayed in varying degree hypermethylation in HCC tissues (Figure 3 and Figure 1). Since hypermethylation of the CAT gene rarely occurred (3.3%) (Figure 3), its impacts on the HCC pathology might be rather trivial. Both GSTpi and CFTR genes were prevalently hypermethylated in HCC and their neighboring non-cancerous tissues (80% and 56.7%, and 77% and 50%, respectively, Figure 1), highlighting the possibility that inactivation of transcription of these two genes may indeed occur in HCC, and have certain etiological significance.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Methylation profiles of the promoter CpG islands of 24 genes in HCC. Both electrophoretic patterns of the representative PCR products of each of 14 targets and the sequencing verification of one representative PCR product were presented. To indicate the methylation status, the sequenced data are aligned with the wild-type sequence. *, size markers, the bands of 250 bp and 100 bp were shown. U, unmethylated; and M, the hypermethylated. Panels: 1, ABCG2; 2, ATF2; 3, B2M; 4, CAT; 5, CFTR; 6, DCK; 7,GSTpi; 8, OCLN; 9, RAF1; 10,RALBP1; 11, SKP2; 12, SPF45; 13, TOP2B; and 14, TP53.

It has been well recognized that the so-called non-cancerous cells pathologically defined may have already suffered from certain genetic lesions as the corresponding cancerous tissues. Inclusion of the neighboring non-cancerous tissues in this study made it possible to analyze our results from the stage-specific perspective of carcinogenesis. As shown in Table 4, no significant difference was detected in the occurrence of hypermethylated GSTpi and CFTR genes between HCC and the paired non-cancerous tissues (P > 0.05) (Figure 1). If the neighboring non-cancerous tissues have indeed suffered from the early stage genetic and/or epigenetic lesions, the methylation changes of these two genes might occur at the early stage of HCC carcinogenesis.

To correlate the hypermethylated status with the transcription silencing of GSTpi gene, we carried out MSP analysis of this gene and semi-quantitative RT-PCR for its expression in five established liver cancer cell lines. As shown in Figure 4, three liver cancer cell lines (i.e. HepG2, Hep3B and L-02) were fully methylated, and did not express GSTpi, while SMMC-7721 and HCCLM3 cell lines were heterozygously methylated, and expressed a detectable level of GSTpi mRNA. Therefore, the general notion stands correct in this case, that the hypermethylated status of the promoter CpG island is inversely correlated with the long-term transcription silencing state of the gene.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Methylation status of the promoter CpG Island correlated the expression state of the GSTpi gene in liver cancer cell lines. Both methylation status (panel 1) and mRNA level (panel 2) of the GSTpi gene in SMMC-7721 (lane 1), Hep3B (lane 2), HepG2 (Lane 3), HCCLM3 (lane 4) are determined and presented. *, size markers, the bands of 250 bp and 100 bp are shown. U, unmethylated; and M, the hypermethylated. The density of the bands (indicated in panel 3) for the intemal control β -actin and the GSTpi gene were densitometrically recorded, where the density of the β -actin in SMMC-7721 cells was arbitrarily taken as 100%. The band-ratio of the GSTpi over the β -actin genes were calculated and summarized in the table and plot (panel 3).

Glutathione-S-transferases (GSTs) are a family of phase II detoxification enzymes that catalyze the conjugation of glutathione to a wide variety of endogenous and exogenous electrophilic compounds, including subclasses of carcinogens and cytotoxic therapeutic drugs[20,21], so that the cells are protected from DNA damages[22]. Furthermore, the GSTpi isoform also regulates the mitogen-activated protein (MAP) kinase pathway that participates in cellular survival and death signals via protein: protein interactions with c-Jun N-terminal kinase 1 (JNK1) and apoptosis signal-regulating kinase (ASK1)[23]. Therefore, GSTs serve two distinct roles in the development of drug resistance to both substrate-type and other types of chemotherapeutic drugs via direct detoxification as well as acting as inhibitors of the MAP kinase pathway[24].

The cystic fibrosis trans-membrane conductance regulator (CFTR) belongs to the ABC transporter superfamily, and regulates the chloride permeability mainly in epithelial cells. Its defects may contribute to cystic fibrosis, a common autosomal recessive disorder among the Caucasian population, which affects multiple organs, including lung, pancreas, liver, sweat gland, reproductive system and intestine. CFTR gene has high expression level of transcripts in epithelial cells in pancreas and nasal polyps; low in lung, liver, and kidney; and undetectable in brain and fibroblasts[25]. It functions as a cAMP-regulated chloride channel, a conductance regulator by an autocrine mechanism involving ATP and/or chloride efflux, and an inhibitor of epithelial Na+ channels when chlorides are released[26]. The involvement of CFTR gene in drug resistance is implicated by observation that the forced over-expression of CFTR gene confers the multiple drug resistance to NIH 3T3 cells[27], a similar behavior to the founding member of this family, multiple drug resistance 1 (MDR1), which was identified three decades ago, for its etiological role in the cellular resistance to the multiple chemotherapeutic drugs. The unmethylated status of CFTR gene in the normal liver tissues suggested that its expression may be necessary for the physiological activities of hepatocytes. Although whether this gene is expressed in HCC has not been tested, the possibility does remain for the functional loss of this gene in HCC, indicated by the prevalent hypermethylation of its promoter CpG islands.

Therefore, lacking of expression of these two genes, implicated by the hypermethylated state of the promoter CpG islands, may make the cells more susceptible to the carcinogenic insults, a feature associated with the proneness to the malignant state of cells. Hypermethylation of the promoter CpG islands may reflect the long-term silencing state of transcription of both GSTpi and CFTR genes in HCC, which may offer growth advantages of the malignant cells over their normal counterparts in vivo. Indeed, the hypermethylation of both genes was an early phase event, as no statistically significant difference in the hypermethylation frequency between the paired neighboring non-cancerous and HCC tissues was observed (Table 4). Nevertheless, the hypermethylated status of these two genes may not be all bad, as the HCC with the hypermethylated alleles of these two genes (if either or both genes did not express) may be more sensitive to the relevant chemotherapeutic drugs than their unmethylated counterparts. Indeed, in a recent report, a better response to chemotherapy was achieved in breast cancer patients lacking GSTpi expression than the expressing counterparts[28].