Get access

Promoter methylation and loss of coding exons of the fragile histidine triad (FHIT) gene in intrahepatic cholangiocarcinomas

Authors


Wolfgang G. Ballhausen, Martin-Luther-University Halle-Wittenberg, KIM1 – Molecular GI Oncology Ernst-Grube-Str. 40, 06120 Halle, Germany.
Tel:+49(0)345 557 5925
Fax:+49(0)345 557 5927
e-mail: wolfgang.ballhausen@medizin.uni-halle.de

Abstract

Abstract: Aims: About 10–30% of primary liver cancers represent intrahepatic cholangiocarcinomas (IHCC). Since chromosomal losses of 3p are detectable in about 40% of cholangiocarcinomas our study aimed at the identification of mechanisms leading to functional deletion of tumor suppressor genes in this region. Our efforts focussed on genomic losses and epigenetic inactivation of two tumor suppressor genes, the fragile histidine triad (FHIT) and the ras association domain family 1 (RASSF1A) genes, both located on the short arm of chromosome 3.

Methods: Methylation-specific PCR (MSP) and combined bisulfite-dependent restriction analysis (COBRA) were applied to detect epigenetic silencing of gene promoters. Genomic duplex PCR was used to identify exon losses of the FHIT gene. Nineteen paraffin-embedded samples of intrahepatic cholangiocarcinomas were studied.

Results: Here we report for the first time that in addition to frequent losses of the exons 5 and 6, hypermethylation of the FHIT promoter occured in a significant portion of IHCC. Methylation specific PCR (MSP) detected epigenetic inactivation of the FHIT/FRA3B locus in 8 of 19 (42%) cases. Combined bisulfite restriction analysis (COBRA) revealed that high levels of methylated FHIT promoter sequences were present in 6 of the 8 methylation positive samples. In agreement with previous reports MSP identified hypermethylation of the RASSF1A gene in 13 of 19 (68%) IHCC specimens examined.

Conclusions: Epigenetic silencing of the FHIT tumor suppressor gene is a novel inactivation mechanism to be considered in the development of intrahepatic cholangiocarcinomas. However, a statistically significant inverse correlation between K-Ras activation and RASSF1A inactivation was not found.

Ancillary