Oxidative stress caused by free radicals has been implicated in the pathogenesis of many cancers.¹ Free radicals, such as reactive oxygen species (ROS), may cause mutations in cancer related genes or directly alter the functions of proteins regulating DNA repair, cell cycle progession, and apoptosis. Oxidative stress has been linked to conditions that are associated with an increased risk of hepatocarcinogenesis, such as viral hepatitis, alcoholic liver disease, toxic liver injury, and fibrosis.^2,3 Furthermore, hepatocellular carcinoma (HCC) shows evidence of oxidative DNA damage, as demonstrated by the formation of 8-hydroxy-2′-deoxyguanosine.^4,5 These observations suggest that ROS induced oxidative damage contributes to the pathogenesis of HCC.

“Mice lacking Prdx1 develop severe haemolytic anaemia and various types of malignancies, including hepatocellular carcinoma, which suggests that this protein functions as a tumour suppressor”

Several studies using karyotyping, comparative genomic hybridisation, and loss of heterozygosity (LOH) analysis have shown that loss of chromosome 1p is a recurrent genetic aberration in human HCC.⁶ Chromosomal losses most often involve the distal regions of chromosome 1p, with the 1p34–36 region involved in up to 75% of HCCs. The shortest regions of overlap on chromosome 1p have recently been mapped to 1p36.22–p36.13 and 1p36.32–p36.22.^7,8 So far, three candidate tumour suppressor genes on the 1p36 region, the PRDM2 (RIZ), RUNX3, and TP73 genes, have been studied in HCC, but no mutations have been detected.^8,9,10

Peroxiredoxin 1 (PRDX1)—also designated MSP23, OSF-3, or PAG—is an intracellular antioxidant protein with thioredoxin dependent peroxidase activity, which protects cells against oxidative stress by inactivating ROS, and is encoded by a single gene on chromosome 1p34.1.¹¹ PRDX1 is highly expressed in the liver and was localised to hepatocytes and Kupffer cells in the rat.^12,13 Mice lacking Prdx1 develop severe haemolytic anaemia and various types of malignancies, including HCC, which suggests that this protein functions as a tumour suppressor.¹⁴ However, PRDX1 gene mutations have not been identified in human tumours.

Because distal chromosome 1p deletions are frequent in human HCC and Prdx1 mutant mice develop HCC with high frequency, we tested the hypothesis that mutations or polymorphisms of the PRDX1 gene may also contribute to the carcinogenesis of human HCC.

MATERIALS AND METHODS

Tumours, DNA extraction

Formalin fixed, paraffin wax embedded tumour samples of 36 HCCs and corresponding non-tumorous liver tissue were drawn from the files of the department of pathology, University Hospital Zürich, Switzerland, covering a period of 11 years (1992 to 2003). Tumour and non-tumorous liver tissue was macrodissected from 10 μm sections and DNA was extracted as described previously.¹⁵

LOH analysis

LOH was investigated using four polymorphic markers of the 1p34.1 region located within 0.8 Mb of the PRDX1 locus (telomeric: D1S421; centromeric: D1S2677, D1S451, and D1S2802). Microsatellite markers were amplified from DNA extracted from HCC and corresponding non-tumorous liver tissue. Standard polymerase chain reaction (PCR) was performed using conditions and fluorescent labelled primers as obtained from the Ensembl Genome Browser (www.ensembl.org). PCR products were analysed on an automated DNA sequencer (model 373A; Applied Biosystems, Foster City, California, USA) using Gene Scan software (Applied Biosystems). Cases were classified as informative when two distinct alleles of similar intensity were found in the normal DNA. LOH was defined as present when an allele peak signal from tumour DNA was reduced by ⩾ 50% compared with the corresponding non-tumorous liver DNA.

PCR, denaturing gradient gel electrophoresis, and DNA sequencing

Primers amplifying the translated exons 2–6 were designed based on the reported genomic sequence of the human PRDX1 gene (ENSG00000117450; www.ensembl.org): exon 2F, 5′-GTA GGT GAA GGC TGC TGG TT-3′; exon 2R, 5′-CAT CTC AAT GAG CCC AGT GTT-3′; exon 3F, 5′-TGC TAA CCA TGA CTC CGA TTT-3′; exon 3R, 5′-TTC ACA TGC CAA ATA GAC CAA-3′; exon 4F, 5′-GAA AGG GGA AAG AGG AAT GC-3′; exon 4R, 5′-GGC TTT CAG CCA ACT GGA TA-3′; exon 5F, 5′-AAC CTG TAT TGC TTT TCC TTT CA-3′; exon 5R, 5′-CAC CCC TTC ATA CCA CCA CT-3′; exon 6F, 5′-ATG GTG GTG CAT ACC ATT GA-3′; and exon 6R, 5′-GCA GCC TGG CAC TAA AAC AG-3′. PCR amplification, denaturing gradient gel electrophoresis, and cycle sequencing of PCR products were performed as described previously.¹⁵ Sequences were compared with the reported genomic sequence of the PRDX1 gene (ENSG00000117450; www.ensembl.org).

RESULTS AND DISCUSSION

Three lines of evidence encouraged us to investigate genetic alterations of the PRDX1 gene in human HCC. First, the PRDX1 gene maps to the distal region of chromosome 1 (1p34.1), a region of frequent LOH in HCC. So far, three candidate tumour suppressor genes at the 1p36 region, the PRDM2 (RIZ), RUNX3, and TP73 genes, have been studied in HCC and no mutations have been detected, leaving the putative tumour suppressor gene(s) in this region unidentified.^8,9,10 Second, HCCs show evidence of oxidative damage, suggesting that defects in ROS scavenging systems may contribute to the initiation and/or promotion of HCC.^4,5 Because the PRDX1 protein participates in the antioxidant defence of hepatocytes, PRDX1 inactivation may enhance oxidative stress and thereby contribute to hepatocarcinogenesis. Third, HCCs occur with high penetrance in Prdx1 mutant mice, suggesting a tumour suppressor function for PRDX1 in hepatocyte progenitor cells.¹⁴

“We cannot exclude the possibility that PRDX1 inactivation by epigenetic mechanisms, such as promoter hypermethylation, may contribute to hepatocarcinogenesis”

To determine whether genetic alterations of the PRDX1 gene are involved in human hepatocarcinogenesis, we first analysed this series of HCCs for LOH of four polymorphic microsatellite markers that are located telomeric and centromeric of the PRDX1 locus on chromosome 1p34. Thirty four of 36 HCCs were informative for at least one of the four markers. Allelic loss of at least one microsatellite marker was seen in three of the 34 informative HCCs. Allelic loss of D1S421, D1S2677, D1S451, and D1S2802 was found in one of 14, two of 21, one of 10, and three of 26 informative cases, respectively (fig 1; table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Loss of heterozygosity (LOH) analysis of a moderately differentiated hepatocellular carcinoma (HCC30) with allelic losses (arrows) of the microsatellite markers D1S421 and D1S2802, which are located telomeric and centromeric of the PRDX1 locus, respectively.

Our results agree with previous studies that also reported LOH at 1p34 in HCC, although with higher frequency (29%,¹⁶ 48–71%,¹⁷ and 15%¹⁸). These discrepancies may result from the use of different microsatellite markers and may indicate that LOH on chromosome 1p occurs distally of 1p34. Mutation analysis by PCR/denaturing gradient gel electrophoresis of the translated exons 2–6 of the PRDX1 gene did not reveal band shift variants in the 36 HCCs. Subsequent DNA sequencing of exons 2–6 of 15 HCCs, including those with LOH at chromosome 1p34.1, revealed wild-type sequences.

We conclude that genetic alterations of the PRDX1 gene are rare in HCC and that PRDX1 is unlikely to be the target tumour suppressor gene of LOH on 1p. However, we cannot exclude the possibility that PRDX1 inactivation by epigenetic mechanisms, such as promoter hypermethylation, may contribute to hepatocarcinogenesis.