Thirty-three samples of HCC and adjacent non-cancerous tissues (> 2 cm away from tumor) were obtained from the Department of Surgery, First Affiliated Hospital of Anhui Medical University between July 2006 and July 2007. These samples were documented through a pathology laboratory database. The age of the 33 patients (27 male, 6 female) ranged between 18 and 67 years, with a median age of 51 years. Ten patients were Edmonson stage I, 18 patients were stage II, and five patients were stage III. The diameter of the tumor was smaller than 3 cm in five patients, and ≥ 5 cm in 28 patients. Portal vein tumor thrombus was found in three patients. As a control, 20 samples of normal liver tissues were collected from the resection of hemangiomas between July 2006 and July 2007. The samples were snap-frozen in liquid nitrogen and stored at -80°C until the extraction of DNA and RNA.

Genomic DNA and total RNA were extracted from HCC tissues, adjacent non-cancerous and normal control tissues using the Qiagen kit (Qiagen). The concentration of DNA and RNA were determined with a spectrophotometer and their integrity was assessed by gel electrophoresis.

We investigated whether the promoter regions of the two genes were methylated. Genomic DNA was modified with sodium bisulfite using the CpGenome DNA Modification kit (Intergen) according to the specifications of the manufacturer. MSP was performed in a total volume of 25 &mgr;L, containing 2 &mgr;L modified template DNA, AmpliTaq Gold (Roche 5 U/&mgr;L) 0.2 &mgr;L, 10 × buffer 2.5 &mgr;L, dNTP (10 &mgr;mol/L) 1 &mgr;L, each primer (10 &mgr;mol/L) 0.5 &mgr;L, Mg²⁺ (25 &mgr;mol/L) 2 &mgr;L, and RNase free water 16.3 &mgr;L. MSP reactions were subjected to an initial incubation at 95°C for 10 min, followed by 35 cycles of 95°C for 45 s, and annealing at the 52-57°C for 45 s and 72°C for 45 s. Final extension was completed by incubation at 72°C for 5 min. Primer sequences of MSP and unmethylation-specific PCR (USP) are shown in Table 1[2]. MSP products were separated on 2% agarose gels and visualized after ethidium bromide staining. The product bands of MSP and USP were calculated using a TANON GIS gel image analysis system, and the relative methylation level was determined by MSP ratio = MSP band density/(MSP band density + USP band density).

Total RNA (1 &mgr;g) was reverse-transcribed with RevertAid First Strand cDNA Synthesis Kit (Fermentas) in a final volume of 20 &mgr;L. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. The cDNAs of interest were quantified by real-time quantitative PCR using the Rotor-Gene 3000. Samples were assayed in a 25 &mgr;L reaction mixture containing 1 &mgr;L of cDNA, 12.5 &mgr;L of 2 × QuantiTect probe PCR Master Mix (Qiagen), 1 &mgr;L of each specific primer, 0.5 &mgr;L of fluorogenic probe and 9 &mgr;L of RNase free water. The sequence of the primer and probe is shown in Table 1. Real-time quantitative RT-PCR reactions were subjected to initial incubation at 95°C for 10 min, followed by 35 cycles of 95°C for 15 s, and 60°C for 60 s.

Statistical analyses were performed using SPSS software version 13.0. χ² test, t test, non-parameter Spearman test and Fisher’s exact test were used to compare the relationship of methylation and clinical data. One-year disease-free survival was analyzed through the log rank test. The level for a statistically significant difference was set at P < 0.05 for all the tests.

Methylation status of DKK-3 and WIF-1 in the three tissues: The methylation status of DKK-3 and WIF-1 in the three kinds of tissues is shown in Table 2 and Figure 1. In HCC and adjacent non-cancerous tissues, the methylation rate of DKK-3 was 37.61% ± 4.26% and 15.96% ± 3.91%, respectively. The methylation status of WIF-1 was 29.70% ± 3.24% in HCC and 18.34% ± 4.02% in adjacent non-cancerous tissues. The methylation rate of DKK-3 and WIF-1 in HCC was higher than that in adjacent non-cancerous tissues (P < 0.05). The methylation of DKK-3 and WIF-1 was not observed in normal control tissues. There was no linear correlation between methylation of DKK-3 and WIF-1 (r = 0.296, P > 0.05).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 DKK-3 (A) and WIF-1 (B) MSP electrophoretogram. T: HCC tissues; N: Adjacent non-cancerous tissues; Z: Control tissues; M: Methylation-specific PCR products bands; U: Unmethylation-specific PCR products bands.

Relationship of methylation and clinical data: The relationship between DKK-3 and WIF-1 methylation and clinical data such as sex, age, Child-Pugh score, AFP, HBsAg, tumor number, cirrhosis, pseudo-capsule and pathology class is shown in Tables 3 and 4. The methylation rate of the DKK-3 gene was higher in older (≥ 60) than younger (< 60) patients, higher in non-cirrhosis than that in cirrhosis patients (P < 0.05), and higher in HBsAg (-) than in HBsAg (+) patients. There was no relationship and no associations observed between the methylation status of DKK-3 and WIF-1 in adjacent non-cancerous tissues and clinical data (P < 0.05).

Relationship of methylation and 1-year disease-free survival: Complete follow-up data was obtained from 27 (27/33) cases of HCC, and the follow-up time was > 1 year. Among the 27 cases, 11 cases showed disease recurrence or died. Kaplan-Meier method and Log-rank test were used to calculate the survival rate and the significance of difference. One-year disease-free survival decreased as time passed, but there was no significant difference in the DKK-3 or WIF-1 between HCC and adjacent non-cancerous tissues (Figure 2).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Relationship between methylation (A) and mRNA expression (B) of DKK-3 gene and 1-year disease-free survival in HCC tissues.

mRNA expression of DKK-3 and WIF-1: The DKK-3 mRNA expression in HCC, adjacent non-cancerous tissues and normal control tissues was 0.773 ± 0.319, 0.833 ± 0.316 and 1.012 ± 0.125, respectively. The expression of DKK-3 mRNA in HCC and adjacent non-cancerous tissues was significantly lower than that in normal control tissues (P < 0.05). There was no significant difference between HCC and adjacent non-cancerous tissues. WIF-1 mRNA expression in the three tissues was 0.853 ± 0.510, 0.820 ± 0.316 and 0.995 ± 0.148, without significant difference among them. Pearson’s product-moment correlation analysis showed that there was no correlation between DKK-3 and WIF-1 in HCC tissues (r = 0.472, P = 0.127).

Relationship between DKK-3 and WIF-1 mRNA expression and clinical data: The relationship between DKK-3 and WIF-1 mRNA expression and clinical data are shown in Tables 5 and 6. DKK-3 mRNA expression in HCC was decreased as the pathological grade increased (P < 0.05), but no significant correlation was found with other clinical data. The expression of DKK-3 mRNA in adjacent non-cancerous tissues, and the expression of WIF-1 mRNA in HCC and adjacent non-cancerous tissues showed no significant correlation with clinical data (P > 0.05). Pearson’s product-moment correlation analysis showed that there was no correlation between DKK-3 and WIF-1 in HCC, adjacent non-cancerous tissues and normal control tissues

(P > 0.05).

Relationship of mRNA expression and 1-year survival: The expression of DKK-3 and WIF-1 gene mRNA in HCC and adjacent non-cancerous tissues was divided into two groups: higher expression group (> average expression) and lower expression group (≤ average expression). Kaplan-Meier curve showed that the disease-free survival was decreased. The survival rate of the lower expression group was lower than that of the higher expression group, but without significant difference (P > 0.05, Figure 2).

Relationship between methylation and mRNA expression: The relationship between mRNA expression and methylation status in HCC and adjacent non-cancerous tissues is shown in Figure 3, and the relationship of DKK-3 and WIF-1 mRNA expression from HCC and adjacent non-cancerous tissues was not found between methylation and non-methylation (P > 0.05).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Relationship between methylation and mRNA expression of DKK-3 and WIF-1 in HCC (A) and adjacent non-cancerous tissues (B). P > 0.05.

The Wnt signal transduction pathway is significantly related to human neoplastic transformation. The Wnt-antagonist genes function as tumor suppressors and contribute to the pathogenesis of several human malignancies. Recently, promoter CpG hypermethylation and gene silencing in DKK-3 and WIF-1 genes have been identified in several human malignancies. The DKK-3 gene locus maps to 11p15.1 of the human chromosome, which encodes a Wnt complex receptor antagonist. Wnt will combine with Frizzle receptor and activate Wnt signaling if the DKK-3 protein is inactivated. The WIF-1 gene locus maps to 12q14.1 of the human chromosome, which encodes a conservatively secreted protein that inhibits the activity of Wnt by holding back regular or non-regular Wnt signal transduction pathway.

Recently, Wnt antagonists have received increasing attention because of their potential role in carcinogenesis. Many studies have reported a relationship between hypermethylation of a gene promoter and low expression of mRNA in several human carcinomas. Promoter hypermethylation and reduced expression of the DKK-3 gene were found in bladder cancer, lung cancer cell lines and tissues.

Few studies have described the correlation between Wnt antagonists and the development of hepatocellular carcinoma (HCC). The results of this study suggest that DKK-3 and WIF-1 may act as negative regulators of Wnt signaling and may be related to the development of HCC. However, this pathogenesis is different from that of hepatitis B virus infection inducing HCC development.

In this study, the authors investigated the methylation status and mRNA expression of DKK-3 and WIF-1 in HCC, adjacent non-cancerous tissues and normal control tissues, and found that the aberrant Wnt antagonists play an important role in carcinogenesis of HCC. This finding may provide a reliable way to improve early diagnosis of HCC and new therapies by blocking this pathway in the treatment of HCC.

MSP: Methylation-specific polymerase chain reaction is a simple, rapid and inexpensive method to determine the methylation status of CpG islands. This approach allows the determination of methylation patterns from very small samples of DNA and can be used in the study of abnormally methylated CpG islands in neoplasia.

Promoter-hypermethylation and the reduced expression of DKK-3 and WIF-1 were shown to be important mechanisms of hepatocarcinogenesis. These results may provide a reliable way to improve early diagnosis of HCC and develop new therapies by blocking this pathway in the treatment of HCC.