Allele loss and epigenetic inactivation of 3p21.3 in malignant liver tumors
Article first published online: 9 FEB 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 115, Issue 5, pages 684–689, 10 July 2005
How to Cite
Tischoff, I., Markwarth, A., Witzigmann, H., Uhlmann, D., Hauss, J., Mirmohammadsadegh, A., Wittekind, C., Hengge, U. R. and Tannapfel, A. (2005), Allele loss and epigenetic inactivation of 3p21.3 in malignant liver tumors. Int. J. Cancer, 115: 684–689. doi: 10.1002/ijc.20944
- Issue published online: 11 MAY 2005
- Article first published online: 9 FEB 2005
- Manuscript Accepted: 22 NOV 2004
- Manuscript Received: 8 SEP 2004
- hepatocellular carcinoma;
- loss of heterozygosity;
Previously, the RASSF1A, BLU and SEMAPHORIN 3B (SEMA3B) candidate tumor suppressor genes on chromosome 3p21.3 were found to be inactivated and downregulated by genetic and epigenetic changes in lung cancer. We analyzed the methylation status of RASSF1A, BLU and SEMA3B in 35 hepatocellular carcinomas (HCCs) and 15 cholangiocarcinomas (CCs) by methylation-specific PCR and loss of heterozygosity (LOH) at 3p21.3 after microdissection. The presence of mRNA transcripts was confirmed by semiquantitative PCR. SEMA3B hypermethylation was found in 29/35 HCCs (83%) and in all (15/15) patients with CC. BLU promoter hypermethylation was detected in 7/35 (20%) HCCs and 3/15 (20%) CCs. In 2 corresponding specimens of hepatitis B virus-related liver cirrhosis, BLU methylation was also observed, but not in uninvolved normal liver tissue. RASSF1A was methylated in 21/35 HCCs (60%) and in 10/15 CCs (67%). LOH at 3p21.3 occurred in 8/35 (23%) HCCs and 3/15 (20%) CCs. The presence of hypermethylation was statistically associated with LOH of SEMA3B and correlated with downregulation of mRNA transcripts. SEMA3B transcripts increased upon treatment of HCC cell lines with the demethylation compound 5-aza-2-deoxycytidine. In conclusion, our data indicate that 2-hit gene silencing of SEMA3B through epigenetic changes and allele loss is a common and important event in the carcinogenesis of malignant liver tumors. © 2005 Wiley-Liss, Inc.
Primary liver cancer is one of the most frequent carcinomas worldwide. Besides hepatocellular carcinoma (HCC), accounting for 80–90% of all primary liver cancers, cholangiocarcinoma (CC) is the second most common primary hepatic malignancy. The short arm of chromosome 3 has been shown to exhibit loss of heterozygosity (LOH) in several types of cancer, including ovarian, kidney, lung, nasopharyngeal and also liver cancer.1, 2 In particular, overlapping homozygous deletions in lung cancers have been identified in region 3p21.3.3, 4 In HCC, LOH of chromosome 3p occurred in about 30% of the patients.5, 6 Several genes located on chromosome 3p have been studied in HCC as well as CC and include RASSF1A on 3p21.31, FHIT at 3p14.2, RIZ1, VHL at 3p25.7, 8, 9, 10 These results directed an intensive search for possible tumor suppressor genes located in the 3p21 region for one or more genes that could function as gatekeepers in molecular pathogenesis of human cancers. A group of candidate tumor suppressor genes (designated BLU, SEMAPHORIN 3B, or RASSF1A) has recently been mapped to this critical gene-rich region.2, 11, 12, 13
SEMAPHORIN 3B (SEMA3B) is a member of the semaphorin family, which is also located on 3p21.3. Semaphorins are a family of signaling molecules initially identified to play a role in axonal guidance and can be classified as either membrane-bound (classes 1, 4, 5 and 6) or secreted (classes 2 and 3).14 The receptors for the class 3 semaphorins are also known as neuropilin receptors. The intracellular signal transduction pathway is less clear but has been shown to involve the Rho family GTPases. Recently, it has been demonstrated that SEMA3B suppresses tumor formation in non small cell lung cancer15, 16 and in an adenocarcinoma model of ovary cancer.17
BLU encoded on 3p21.3 was found to be rarely inactivated in lung cancer, although expression was downregulated in a subset of lung tumor cell lines. BLU promoter region hypermethylation was observed in cervical cancer,18 lung, breast, kidney, nasopharyngeal and neuroblastoma cell lines.19 This study was performed to analyze the status of RASSF1A, SEMA3B and BLU in malignant liver tumors.
Material and methods
Patients and tissue samples
Thirty-five patients with HCC and 15 with CC undergoing partial hepatectomy (segmental or lobar resection) between 1996 and 2000 (38 male, 12 female; median age, 68; range, 49–71) were included in this retrospective study. Tumor samples were compared with unaffected specimens of the same individuals. No patient received preoperative or adjuvant chemo- or radiotherapy. All patients underwent surgery with curative intent (R0 resections). Patients who received orthotopic liver transplantation were excluded from this study. Tumor typing and staging were performed using WHO20 and UICC21 criteria, respectively. In 16/35 (46%) patients with HCC, a liver cirrhosis was present. Two patients with liver cirrhosis were HBV-positive, 12 cases had an alcoholic cirrhosis, whereas in 2 patients the etiology of the cirrhosis remains unclear. In case of CC, no concomitant liver cirrhosis was observed.
All patients were informed of special examination of tumor samples, which was in accordance with the ethical standards of the Committee on Human Experimentation of the University of Leipzig. All samples were taken during treatment procedures with therapeutic intent.
Allelic losses were analyzed by a PCR approach with primers amplifying polymorphic microsatellites flanking the 3p21.3 gene at loci D3S1568, D3S1621 and D3S4597. The primer sequences were obtained from the Genome Database and were amplified by PCR containing [1α-32P]dCTP (DuPont New England Life Science Products). PCR was performed on the genomic DNA samples for PCR cycles, including one cycle of 95°C for 12 min followed by 35 cycles consisting of 10 cycles at 94°C for 15 sec, 55°C for 15 sec and 72°C for 30 sec and 25 cycles at 89°C for 15 sec, 55°C for 15 sec and 72°C for 30 sec, followed by 72°C for 10 min. PCR product (2 μl) was denatured in formamide and separated by electrophoresis on a 6% polyacrylamide-7 M urea gel at 60 W at room temperature. PCR products were detected by autoradiography (Biomax film; Eastman Kodak, Rochester, NY). LOH was visually scored by > 50% reduction in allele intensity compared with the normal counterpart.
After microdissection,22, 23 methylation-specific PCR (MSP) was used to investigate the methylation status of the promoter region of the BLU, RASSF1A and SEMA3B genes. Microdissection was performed as described previously.22, 23 After an initial bisulfite treatment to modify the DNA, PCR was performed to distinguish methylated from unmethylated DNA as described by Herman et al.24 Briefly, 2 μg of genomic DNA was denatured with 0.3 M NaOH; 10 mM hydroquinone and 3 M sodium bisulfite were added and incubated for 16 hr at 50°C. Afterward, modified DNA was purified using the Wizard DNA purification resin (Promega, Madison, WI) followed by desulfonating in 0.3 M NaOH and subsequent ethanol precipitation and resuspension in 30–50 μl water. MSP was performed using specific primers and conditions already described.2, 3, 8, 15, 19, 24 Detailed primer sequences are available on request. Briefly, a 20 μl reaction volume contained 150 ng bisulfite-modified DNA, 1 × PCR buffer, 1.5 mM MgCl2, 0.16 μM dNTPs, 0.25 μM specific primer mix (forward and reverse primers) and 1 unit Taq enzyme (Roche, Mannheim, Germany). The BLU-specific reaction mixture was incubated at 95°C for 10 min with a hot start, followed by 6 touchdown cycles at 95°C for 20 sec, 66°C for 40 sec, 72°C for 20 sec, X − 1°C per cycle, and then 30 cycles of 95°C for 20 sec, X − 6°C for 30 sec, 72°C for 30 sec, 72°C for 10 min (UmspX = 66°C; mspX = 72°C). The SEMA3B cycling conditions were 95°C for 10 min with a hot start, followed by 40 cycles at 95°C for 30 sec, 58°C for 30 sec, 72°C for 30 sec and then 72°C for 10 min. For RASSF1A, a hot start PCR was performed at 95°C for 10 min, followed by 29 cycles of 95°C for 30 sec, 60°C for 30 sec, 72°C for 30 sec and then 72°C for 10 min, adopted from the literature.8
The PCR products were electrophoresed on a 2% agarose gel, stained with ethidium bromide and visualized under UV illumination. Placental DNA treated with methyltransferase was used as a positive control for methylation. Furthermore, the cell lines HepG2 and TFK-1 were analyzed as positive control for RASSF1A methylation, TFK-1 for Blu, as well as Sema3B methylation.
The presence of SEMA3B, BLU and RASSF1A mRNA transcripts was analyzed by semiquantitative PCR (LightCycler; Roche). A total of 500 ng of RNA extracted from 30 mg tissue sample using the RNeasy Mini kit (Qiagen, Hilden, Germany) was reverse-transcribed with SEMA3B (forward, 5′-TTCTTTCGTGAGACGGCGGTA-3′; reverse, 5′-CCCTGGAAGATGCTGCTGGA-3′) and BLU (forward exon 6.2, 5′-GCTTAGCACACACAACCTGC-3′; reverse exon 8, 5′-CCTTGGCAAAACTTGTGAGG-3′) according to the previously published protocols15, 19 and for RASSF1A-specific primers (forward, 5′-TCCTGCAAGGAGGGTGGCTTC-3′; reverse, 5′-GGCTGGGAACCCGCGGTG-3′) in 20 μl of RT-PCR-specific reaction mix with QuantiTect SYBR Green RT-PCR kit (Qiagen) in accordance with the manufacturer's instructions. The size of the PCR products obtained was 274 bp for SEMA3B, 233 bp for BLU and 238 bp for RASSF1A. β-actin was used to provide an internal marker for mRNA integrity (forward, 5′-TCACCATGGATGATGATATCGC-3′; reverse, 5′-AAGCCGGCCTTGCACAT-3′; product size, 66 bp).
To induce expression of SEMA3B after exposure to 5-aza-2-deoxycytidine (5-AZA-C), a drug that inhibits DNA methylation, subconfluent cultures of the SEMA3B-nonexpressing cell line (HepG2) were exposed to 1 μM 5-AZA-C for 7 days. After isolation of total RNA using RNeasy extraction kit (Qiagen), RT-PCR was performed for SEMA3B as described above.
Associations between methylation status, LOH and mRNA transcripts as well as clinicopathologic features were calculated using chi-square statistics. For all calculations, SPSS statistical software (SPSS, Chicago, IL) was used. The significance level was defined as p < 0.05.
To analyze the methylation status of the promoter region of SEMA3B in HCC as well as CC, MSP analysis was used. Hypermethylation of the promoter region of SEMA3B was detected in 29 out of 35 HCCs (83%; Fig. 1a) and in all 15 cases of CC (100%; Fig. 1b, Table I). Unmethylated bands were detected in most tumor samples (Fig. 1). Unmethylated bands were detected in all of the corresponding noncancerous tissues (i.e., normal and cirrhotic liver, irrespective of etiology). There were no significant correlations of SEMA3B promoter hypermethylation with patients' age, stage, or grade of the tumor and prognosis.
|Number of patients||1-year-survival rate, % (95% CI)||Median survival time, days (95% CI)|
|Methylated||29 (83%)||48 (30–86)||391 (0–940)|
|Unmethylated||6 (17%)||51 (36–92)||423 (0–720)|
|Methylated||15 (100%)||81 (37–100)||381 (156–539)|
|Methylated||7 (20%)||53 (32–65)||389 (0–610)|
|Unmethylated||28 (80%)||42 (19–71)||440 (0–940)|
|Methylated||3 (20%)||89 (80–100)||391 (300–539)|
|Unmethylated||12 (80%)||71 (37–100)||371 (126–510)|
|Methylated||21 (60%)||50 (29–61)||375 (0–642)|
|Unmethylated||14 (40%)||51 (19–69)||381 (0–850)|
|Methylated||10 (67%)||67 (50–100)||381 (0–541)|
|Unmethylated||5 (33%)||85 (64–95)||401 (8–980)|
BLU promoter hypermethylation was detected in 7/35 (20%) HCCs (Fig. 2a) and 20% (3/15) CCs (Table I). Importantly, in all patients, corresponding nonneoplastic liver tissue was also analyzed. In 2 cases of liver cirrhosis attributable to chronic hepatitis B virus infection, BLU methylation was also observed (Fig. 2). In case of multiple tumor nodules, all nodules depicted an identical BLU status (Fig. 2b, depicted for tumor nodule a and b).
No correlation of the status of BLU methylation with regard to tumor stage, grade, or prognosis was detected. RASSF1A promoter methylation was detected in 21 out of 35 (60%) HCCs and in 10 out of 15 CCs (67%; data not shown). Hypermethylation of RASSF1A was also detected in 8 cases of concomitant liver cirrhosis in patients with HCC. All 15 noncirrhotic liver specimen of CC patients exhibited an unmethylated RASSF1A promoter. Despite microdissection, amplification of unmethylated templates of SEMA3B, BLU and also RASSF1A was also detected to some degree, probably because of admixed normal intratumorous cells (fibroblasts, endothelial cells, inflammatory cells).
To analyze the expression of SEMA3B and BLU on the transcript level, LightCycler PCR was performed. SEMA3B expression was examined in all 35 HCC and 15 CC specimen as well as in corresponding nontumorous liver tissue (Fig. 3). RT-PCR demonstrated an absence (tumor 1) or reduction (tumor 36) of SEMA3B transcripts in 29 HCC and in all cases of CC tissues (data not shown). All these tumors exhibited a methylated SEMA3B promoter. The remaining 6 HCC specimens as well as all noncancerous tissues showed SEMA3B mRNA expression (as indicated with “N” for normal liver in Fig. 3a). RT-PCR revealed that all samples with an unmethylated BLU promoter expressed corresponding mRNA at a level comparable to controls (normal nonneoplastic liver with or without cirrhosis; Fig. 3a and b; indicated with “N”). BLU mRNA was either reduced (tumor 14) or absent (tumor 10) in those specimens with promoter methylation (Fig. 3b and d). In case of RASSF1A, promoter methylation was also accompanied with a downregulated mRNA transcript.
To confirm that promoter hypermethylation contributes to the lack of expression of SEMA3B in HepG2 cell lines, the effect of 5-AZA-C, a drug that inhibits DNA methylation, was analyzed. After exposure of the SEMA3B-nonexpressing cell line HepG2 to 5-AZA-C for 3 days, reexpression of SEMA3B was detected (Fig. 4), with little or no change in the expression of the housekeeping gene β-actin being observed.
LOH at 3p21.3 was detected in 8/35 (23%) HCCs and in 3/15 (20%) CCs that were informative for at least 2 of the 3 markers at 3p21.3 (Fig. 5). We also examined the relationship between methylation status and the presence of LOH. Among 8 informative HCC cases for the SEMA3B region, 7 of 8 tumors (88%) with LOH showed hypermethylation of SEMA3B, while 2 of 27 tumors without LOH showed hypermethylation of SEMA3B. The correlation between SEMA3B hypermethylation and LOH at the SEMA3B region was statistically significant (Fisher's exact test, p = 0.01). In CCs, all 3 cases with LOH at 3p21.3 exhibited also a hypermethylated SEMA3B promoter. Out of 21 HCC specimens with a methylated RASSF1A promoter, 2 exhibited LOH on 3p21. All 3 CCs with LOH on 3p21 had an unmethylated RASSF1A promoter. Neither in HCCs nor in CCs was hypermethylation of the BLU promoter accompanied with allelic retention.
In this study, we demonstrated that SEMA3B hypermethylation significantly correlated with LOH of SEMA3B locus at 3p21.3 and also with the absence of SEMA3B mRNA expression in HCCs as well as CCs. Previous studies have demonstrated that mutations in the SEMA3B gene are infrequent events. Sekido et al.2 and Kuroki et al.15 reported that loss of SEMA3B mRNA expression was detected in approximately 50% of non small cell lung cancer, which was caused by SEMA3B promoter hypermethylation. Conversely, the tumor suppressor effect could be restored by adenovirus-mediated expression of genes located on chromosome 3p.21.3.25 It has also been shown that treatment with the demethylating drug 5-AZA-C restored SEMA3B expression in these cell lines, suggesting that SEMA3B promoter hypermethylation is responsible for silencing SEMA3B expression.26
We could show in this study that SEMA3B methylation together with LOH on 3p21 occurred in most cases of HCC and in all patients with CC not only in advanced-stage tumors but also in early stages and in different nodules of a given tumor. These findings suggest that SEMA3B promoter hypermethylation is a relatively early-stage event in the pathogenesis of HCC and CC, potentially facilitated by hepatitis B or C virus infection, as has been described for other promoters in HCC.27 Further studies, especially on precursor lesions (e.g., dysplastic nodules), are necessary to answer these question.
Our findings suggest that SEMA3B gene alterations may play a role in the tumorigenesis of HCC and CC via a 2-hit mechanism, including epigenetic changes and allelic loss of tumor suppressor genes. To a lesser extent, BLU and RASSF1A, the additional tumor suppressor candidates on chromosome 3p21, were also inactivated in a number of HCCs and CCs.
Interestingly, promoter methylation of BLU and RASSF1A was also detected in 2 cases of liver cirrhosis, as has been previously reported for HCC.8 This was in accordance to Liu et al.18 and Schagdarsurengin et al.,8 who found promoter methylation of CpG islands of the BLU and RASSF1A genes in normal epithelium, blood lymphocytes and liver cirrhosis related to inflammation or increasing age of the patient. Recently, it has been shown that RASSF1A regulates progression of mitosis by inhibiting the APC-Cdc20 complex and the stability of mitotic cyclins.28 Therefore, loss of RASSF1A may contribute to tumor progression by inducing both disturbances of mitotic progression and chromosome instability.28
In conclusion, 2-hit gene silencing of SEMA3B through epigenetic changes and allelic loss is a common and important event in the carcinogenesis of malignant liver tumors. Our data suggest elucidating the mechanistic role of SEMA3B as well as BLU and RASSF1A in the pathogenesis of liver cancers in more detail.
- 11The 630-kb lung cancer homozygous deletion region on human chromosome 3p21.3: identification and evaluation of the resident candidate tumor suppressor genes—the International Lung Cancer Chromosome 3p21.3 Tumor Suppressor Gene Consortium. Cancer Res 2000; 60: 6116–33., .
- 20WHO: pathology and genetics. Tumours of the digestive system. In: HamiltonSR, AaltonenLA, eds. Lyon: IARC, 2000. 173–80.
- 21UICC: TNM classification of malignant tumors. In: SobinLH, WittekindC, eds. 6th ed. New York: Wiley-Liss, 2002.