Carcinogenesis is a multistep process that involves the accumulation of mutations in genes that govern cell proliferation, regeneration and apoptosis.1, 2 Although the molecular events directly linked to the development and progression of human hepatocellular carcinoma (HCC) are not well understood, it is believed that a step-by-step process of genetic alterations accounts for hepatocarcinogenesis.3, 4 Epidemiological studies have demonstrated that most HCCs arise in the setting of chronic liver disease with the features of chronic hepatitis (CH) or liver cirrhosis (LC). It has been suggested that the increased rate of cell regeneration observed in CH and LC might result in the enhanced susceptibility of hepatocytes to genetic changes; however, the molecular mechanism of how cancer cells acquire a number of genetic changes has not yet been identified.5 One mechanism for such enhanced susceptibility is the impairment of the mismatch repair system.6, 7, 8, 9 However, the frequency of such defects in the DNA repair system is generally low in human cancers, and in fact, the involvement of microsatellite instability has been shown to be rare in viral hepatitis–related hepatocarcinogenesis.10, 11 Accordingly, the question still remains as to how a large number of somatic mutations occur in human cancers including HCC.
Antigen stimulation of mature B lymphocytes induces somatic hypermutation, which diversifies the variable region of immunoglobulin genes. Somatic hypermutation requires activation-induced cytidine deaminase (AID), an enzyme that is expressed by activated B cells in germinal centers.12, 13 Recent studies have demonstrated that inappropriate expression of AID could act as a DNA mutator that contributes to tumorigenesis.14 In fact, a number of studies have shown a high expression of AID transcripts in human lymphoid malignancies, including non-Hodgkin's lymphomas and chronic lymphocytic leukemia.15, 16, 17, 18 We previously demonstrated that aberrant AID expression serves as a link between the cellular editing machinery and a high mutation frequency leading to lymphoid and nonlymphoid malignancies, using AID transgenic (Tg) mice model.19 Most AID Tg mice developed microadenomas of the lung epithelium as well as T-cell lymphomas exhibiting frequent point mutations in the T-cell receptor and c-myc genes that appeared to be introduced because of AID activity.19 To further define the oncogenic impact of aberrant AID expression on nonlymphoid tissues, we recently analyzed the liver phenotype of AID Tg mice. Notably, we found that AID Tg mice also developed liver tumors with the morphological appearance of HCC (unpublished data). These findings prompted us to speculate that ectopic expression of AID could lead to tumorigenesis in liver tissues as well.
In this study, to examine the involvement of AID in the development of human HCC, we examined the AID expression and its association with mutation frequencies of the p53 gene in human normal liver tissue, tumors and surrounding noncancerous liver tissues with underlying CH or LC from patients with HCC. The inducibility of AID expression by cytokine stimulation was analyzed in cultured human hepatocytes.
AID, activation-induced cytidine deaminase; APOBEC-1, apolipoprotein B mRNA catalytic subunit 1; CH, chronic hepatitis; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; LC, liver cirrhosis; OHT, 4-hydroxytamoxifen; RGYW/WRCY, [R, purine (A/G); Y, pyrimidine (C/T); W, A/T]; RT-PCR, reverse transcription polymerase chain reaction; SD, standard deviation; SE, standard error; Tg, transgenic; TGF-β, transforming growth factor β; 18s rRNA, 18s ribosomal RNA.
Material and methods
The study group consisted of 51 patients who had undergone potentially curative resection of primary HCC at Kyoto University Hospital from 2000 to 2003. Selection of patients enrolled in this study was based on the availability of a sufficient amount of tissue for analysis. The patients included 34 men and 17 women, with an average age at the time of surgery of 63 ± 11 years (mean ± SD; range, 39–81 years). Serum samples were tested for serological markers of hepatitis B virus (HBV) and hepatitis C virus (HCV) using commercial enzyme immunoassay kits (Dainabot, Tokyo, Japan) as well as HCV-RNA (Roche Co., Tokyo, Japan).20 Fourteen patients were positive for hepatitis B surface antigen (HBsAg), and 30 were positive for serum antihepatitis C antibody (anti-HCV). Based on the blood chemistry, ultrasonography and histopathological findings, the enrolled patients were classified into 2 groups: those with CH (n = 25) and those with the features of LC (n = 26). As a control, 6 samples of normal liver tissue from patients with metastatic liver cancer were also examined. Biopsies of tumor tissues and nontumor tissues at the proximal edge of freshly resected specimens were obtained and immediately frozen in liquid nitrogen. Total RNA was extracted from tissue specimens using the guanidinium-phenol-chloroform method (Sepasol; Nacalai Tesque, Kyoto, Japan).20 Written informed consent for the use of their resected tissues was obtained from all patients in accordance with the Declaration of Helsinki, and the Kyoto University Graduate School and Faculty of Medicine Ethics Committee approved the study.
Reverse transcription was performed using the Superscript III first strand synthesis system and oligo-dT12–18 primers (Invitrogen, Carlsbad, CA) as described elsewhere.20 The oligonucleotide primers used for human AID amplification were 5′-AAATGTCCGCTGGGCTAAGG-3′ (forward) and 5′-GGAGGAAGAGCAATTCCACGT-3′ (reverse). For the selective detection of endogenous AID transcript, 2 oligonucleotide primers were synthesized as follows: 5′-TATTTTTACTGCTGGAATACTTT-3′ (forward) and 5′-TGACATTCCTGGAAGTTGCTA-3′ (reverse). This primer set amplifies the AID sequences including the 3′-untranslated region; thus the sequences of AID derived from the AID-ER expression plasmid (described later) were not expected to be amplified using these primers. The oligonucleotide primers used for APOBEC-1 amplification were 5′-GAGGCCTGTCTGCTCTACGAA-3′ (forward) and 5′-CTTCCACGTGATTGGTGGTG-3′ (reverse).
The quantification of gene expression was performed by quantitative real-time reverse transcription polymerase chain reaction (RT-PCR) using an iCycler-iQ apparatus and iQ Supermix according to the manufacturer's instructions (Bio-Rad, Hercules, CA). The 6-carboxyfluorescein (FAM)-labeled probe used for human AID was 5′-TCGGCGTGAGACCTACCTGTGCTAC-3′. Standard curves for AID were generated for every target using a 10-fold serial dilution series of 5 independent transcripts derived from BL2-lymphoma cells that contained a high expression of endogenous AID.
To assess the quantity of isolated RNA, as well as the efficiency of cDNA synthesis, target cDNAs were normalized to the endogenous mRNA levels of the housekeeping reference gene 18s ribosomal RNA (18s rRNA). The oligonucleotide primers and FAM-labeled probe used for 18s rRNA amplification were 5′-TAGAGTGTTCAAAGCAGGCCC-3′ (forward), 5′-CCAACAAAATAGAACCGCGGT-3′ (reverse) and 5′-CGCCTGGATACCGCAGCTAGGAATAATG-3′ (probe). For simplicity, the ratios were represented as relative values when compared to expression levels in a lysate from BL2-lymphoma cells. The reproducibility of this quantification method was examined by comparing the results obtained from replicate samples during the same reaction run and those from independent runs on different days.21 The PCR procedures were performed at least 3 times for each sample.
Human hepatoma-derived cells (PLC/PRF/5, HepG2, Hep3B, Huh6 and Huh7) were cultured at 37°C in Dulbecco's modified Eagle's medium (Gibco-BRL, Tokyo, Japan) supplemented with 10% fetal bovine serum.22 Human B cell lymphoma derived cells, FL218 and Tree92, were cultured at 37°C in RPMI-1640 medium (Gibco-BRL).
To analyze the ability of AID to induce somatic mutations in hepatocytes, we established a system that allows a conditional expression of active form of AID. HepG2 cells were transfected with ScaI-linearized pAID-ER-BOSbsr vector encoding active form of AID fused with the hormone-binding domain of the human estrogen receptor (ER) designed as AID-ER.23 After 2 days, cells were cultured in a medium containing Blasticidin S HCl (Invitrogen) until colonies of stably transfected clones arose.
Establishment of human primary hepatocytes in culture
Normal liver tissue was obtained from the noncancerous tissues of the patients with metastatic liver cancer, and hepatitis virus-infected liver tissue from those who had HBV-induced active hepatitis at the time of operation. The isolation and culture of primary human hepatocytes was carried out according to a previously described protocol.24 Briefly, the liver specimens were infused through vascular cut ends using a syringe with warmed (37°C) Hanks' solution containing 0.5 mM EDTA (Sigma, St.Louis, MO) and 1 M HEPES (Wako Pure Chemical Industries, Osaka, Japan). The tissue samples were then subjected to enzyme digestion using 0.05% collagenase (Wako). The cell suspension was passed through double layers of sterile cotton gauze and filtered through a single 75 μm mesh layer of nylon cloth. The cells were then centrifuged at 50g for 2 min to collect the mature hepatocytes. This procedure was repeated 3 times. The liver-infiltrating lymphocytes were isolated from the supernatant on a Ficoll gradient centrifugation according to a standard procedure (Wako). The hepatocytes obtained were suspended at a density of 5 × 105 cells/ml in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. The normal hepatocytes were further treated with human transforming growth factor β (TGF-β; Peprotech, London, UK) after 24 hr of cell seeding to allow attachment to the dishes. Total RNA was extracted from those primary cells 12 hr post-TGF-β treatment.
Immunoblotting and immunohistochemistry
Homogenates of liver specimens were diluted in 2× sodium dodecyl sulfate (SDS) sample buffer (62.5 mM Tris-HCl, pH 6.8; 2% SDS; 5% β-mercaptoethanol; 10% glycerol and 0.002% bromophenol blue) and boiled for 3 min. Protein samples were separated by SDS-polyacrylamide gel electrophoresis (PAGE) on 12% (w/v) polyacrylamide gels and subjected to immunoblotting analyses.25 The polyclonal antibodies against human AID used in this study have been described previously.26 Anti-α-tubulin antibody was obtained from Calbiochem (CA, USA). Immunohistochemistry was carried out according to a previously described protocol.27
Subcloning and sequencing of the human p53 gene
The oligonucleotide primers for human p53 were designed to amplify the sequences between exon 2 and exon 11 spanning 7040 bp of p53 gene using the following primers: p53-182S 5′-GCCGAATTCATTGGCAGCCAGACTGCCTTC-3′, p53-1423AS 5′-ATCCTCGAGTCAGTGGGGAACAAGAAGTG-3′, p53-522S 5′-ACCAGGGCAGCTACGGTTTC-3′ and p53-817S 5′-GCGTGTGGAGTATTTGGATG-3′. Amplification of the p53 gene was carried out using the Expand High Fidelity PCR system (Roche, Mannheim, Germany) and the products were subcloned by insertion into the pcDNA3 vector (Invitrogen) as described previously.22 The resulting plasmids were subjected to sequence analysis using a DYEnamic ET terminator kit with AmpliTaq DNA polymerase (Amersham Pharmacia Biotech, Piscatway, NJ) on an automated sequencer (Applied Biosystems).
Statistical differences in AID expression levels were analyzed using the Mann–Whitney U test. Values of p < 0.05 were considered statistically significant.
Expression of AID in human liver tissues
To examine the AID expression in the human hepatocytes under physiological or pathological conditions, we quantified the AID transcripts in human liver tissues. We found that AID was transcribed only in trace amounts in normal liver tissues, with AID/18s rRNA ratios averaging 7.6 ± 2.5 (mean ± SE). However, 41 (80%) of the 51 cancerous tissue samples showed AID expression above the average level in normal liver (Fig. 1a). The mean AID/18s rRNA ratio in tumor tissues (78.1 ± 21.0) was significantly higher than that in normal liver (p < 0.05). The average AID expression in the noncancerous tissues that exhibited the histological features of CH or LC was 81.2 ± 18.9, also significantly higher than that in normal liver (p < 0.05). A comparison of paired nontumor and tumor samples showed that mRNA concentrations of AID were higher in 26 of 51 (51%) tumors, while the remaining cases (49%) showed a higher expression of AID in nontumor than in tumor tissues. Thus, there was no significant difference of AID expression levels between tumor tissue and surrounding noncancerous tissues with chronic liver disease.
We then compared the AID expression levels in cirrhotic liver specimens with those in CH tissues (Fig. 1b). The mean AID mRNA concentrations tended to be higher in the cirrhotic livers (122.2 ± 34.3) than in the CH livers (38.7 ± 10.0), approaching statistical significance (p = 0.06).
As chronic liver damage can be of various etiologies, we tested whether AID is specifically overexpressed in livers with specific hepatitis viral infections. Both the tumor and nontumor liver tissues with HCV infection exhibited significantly higher expression levels of AID gene when compared with liver tissues of normal control (p < 0.01), whereas there was no significant difference in AID gene expression levels between the tumor as well as nontumor liver tissues with HBV infection and tissues of normal control (Fig. 1c). We also investigated the association of AID mRNA levels with various kinds of clinicopathological factors; however, none of the parameters examined correlated significantly with AID expression (data not shown).
Expression of AID transcripts in human hepatocytes underlying viral hepatitis
AID expression was significantly increased in the liver tissue accompanied with chronic inflammation. To determine whether the increased expression of AID is derived from hepatocytes or infiltrating B lymphocytes, we separated the cells from the nontumor tissues of patients with HBV-induced active hepatitis into hepatocyte- and liver-infiltrating lymphocyte-fractions, and quantified the expression levels of AID in each fraction. First, we observed that albumin and CD19 mRNA were specifically amplified in each fraction, indicating that those fractions selectively consisted of hepatocytes and lymphocytes, respectively (Fig. 2). Then, AID transcripts were amplified using quantitative real-time RT-PCR with cDNA from each fraction as the template. We found a considerable level of AID in both hepatocytes and lymphocytes, indicating that hepatocytes derived from the liver with underlying hepatitis-virus–induced inflammation exhibit a high expression of AID. Similar findings were obtained using the liver tissues from the cases with HCV infection (data not shown).
Expression of AID protein in human liver tissues
To study the expression of AID protein, we performed immunoblot analyses on liver tissues from normal and HCC specimens using antibodies specific for human AID. Expression of AID protein was not detected in the lysates from normal liver tissues, consistent with the RNA data obtained from the quantitative real-time RT-PCR analyses. In contrast, AID protein expression was clearly elevated in most of the HCC and surrounding liver tissues underlying CH or LC, though AID protein levels varied among the specimens examined (Fig. 3a).
To clarify the specific expression and precise localization of AID protein in human liver tissues underlying chronic liver disease, the expression of AID protein was examined by immunohistochemistry in various paraffin-embedded liver specimens. The specificity of the immunostaining results was confirmed by control staining performed on germinal centers of lymphoid organs, which contain mostly activated B cells (Fig. 3b-a).12 In cirrhotic liver tissues, immunoreactivity for AID was indeed detectable in hepatocytes as well as some of the lymphocytes infiltrating around portal and periportal area (Figs. 3b-c and 3b-d). In HCC tissues, AID protein expression was observed in neoplastic cells, localizing mainly in the cytoplasm (Figs. 3b-e and 3b-f). In contrast, no immunostaining of AID was observed in the normal liver tissues (Fig. 3b-b).
AID expression and mutation frequencies in the p53 gene in the liver underlying chronic hepatitis or liver cirrhosis
Animal model have shown that the aberrant expression of AID contributes to tumorigenesis concomitant with frequent mutations in various genes. In the light of these findings, we examined the relationship between somatic mutation frequencies and AID expression levels in liver tissues with AID upregulation. In this study, we focused on the p53 gene, as downregulation and somatic mutations in this gene have been well characterized in human hepatocarcinogenesis.5 To investigate the overall mutation frequency, we determined the sequences between exon 2 and exon 11 of the p53 gene of 63 randomly picked clones that were amplified from tumor and nontumor tissues from 5 HCC patients (Table I). Among them, 4 had HCV infection and the remaining patients lacked any evidence of hepatitis virus infection. We first confirmed that less than 1 substitution out of 1 × 104 nucleotides was detected in the p53 genes subcloned from the normal liver (data not shown). In contrast, a number of nucleotide alterations were observed in the p53 gene of both tumor and nontumor tissues of HCC patients with endogenous AID expression. Some of the nucleotide alterations were clustered in exon 4 of the p53 gene. Taken together, these findings indicate that p53 gene somatic mutations concomitant with aberrant expression of AID are present in the liver tissues underlying CH or LC.
Table I. AID Expression and Mutational Profile of the p53 Gene in Humanhepatocellular Carcinomas
HCV indicates hepatitis C virus infection; NonB NonC, tissues lacking evidence of hepatitis virus infection.
Relative AID levels are shown as AID/18s rRNA mRNA ratio.
Number of mutated clones/total number of clones examined.
Mutation frequency of p53 genes is expressed as the number of mutated nucleotide per 1 × 104 nucleotides in the p53 gene.
Case 1 (HCV)
Case 2 (HCV)
Case 3 (HCV)
Case 4 (HCV)
Case 5 (NonB NonC)
TGF-β-mediated AID expression in human hepatocytes
Our findings that AID expression was specifically upregulated in hepatocytes and neoplastic cells underlying chronic inflammation prompted us to examine whether certain cytokine stimulation contributed to AID expression in those cells. To test this hypothesis, 5 hepatoma cell lines were analyzed by conventional RT-PCR analyses. We found that AID was transcribed at substantial levels in most of the hepatoma-derived cells (Fig. 4a). Although the expression of AID was low in HepG2 cells, marked upregulation of AID transcripts was observed after treatment with TGF-β, a cytokine that can induce AID in B cells (Fig. 4a).12 To further investigate whether AID was induced in a TGF-β-dependent manner, we examined the expression of AID protein in HepG2 cells after the stimulation with TGF-β. The expression of AID protein was substantially increased in response to TGF-β treatment in HepG2 cells (Fig. 4b).
To examine whether TGF-β generally enhances AID expression in hepatocytes, we analyzed AID expression in human hepatoma-derived PLC/PRF/5 cells. Marked upregulation of AID transcript was observed after treatment with TGF-β in PLC/PRF/5 cells, whereas the expression of APOBEC-1 and β-actin was unchanged (Fig. 4c). Quantitative real-time RT-PCR analysis with FAM-labeled probes revealed that TGF-β treatment induced a dose-dependent increase in AID expression with a maximum level at 4 ng/ml, followed by a decrease at 8 ng/ml (Fig. 4d). Moreover, TGF-β induced a time-dependent transcriptional upregulation of AID with its peak level at 12–15 hr after the treatment, whereas the expression of 18s rRNA transcripts was unchanged (Fig. 4e).
To confirm the TGF-β-mediated induction of AID expression in normal human hepatocytes, we established cultured primary human hepatocytes from normal liver tissues of patients with metastatic liver cancer, and investigated whether AID expression is induced by TGF-β. Endogenous AID transcripts were undetectable by RT-PCR analysis in human primary hepatocytes in the resting state (Fig. 4f). However, AID expression was substantially induced after treatment with TGF-β, whereas the expression of albumin transcripts was unchanged.
Somatic mutations in the p53 gene in the cells with AID activation
To obtain the evidence showing that aberrant AID expression could contribute to the abnormal editing of the p53 genes, we established the system that allows the conditional AID activation in cells by introducing an estrogen analogue, 4-hydroxytamoxifen (OHT). In response to OHT, AID is promptly activated by posttranslational conformational change in AID-ER expressing HepG2 cells. Using this system, we investigated the mutation frequencies in the p53 genes. HepG2 cells were stably transfected with either control or AID-ER-encoding plasmids. RT-PCR analysis of the resulting cell lines demonstrated overexpression of AID transcript in the cells harboring the AID-ER-encoding plasmid but not in the cells with the control vector, while only trace amounts of endogenous AID were detectable in both cells (Fig. 5a). Transfectants of AID-ER were then treated with OHT for 13 days and the nucleotide sequences of the p53 gene of randomly picked clones were determined in the cells with and without OHT treatment. In the control cells, no nucleotide mutation emerged by introducing OHT. In contrast, several nucleotide alterations appeared in the whole coding region of the p53 gene in HepG2 cells after AID activation (Fig. 5b). The mutation frequency observed in the cells with AID activation was estimated as 1.4 ×10−4 mutations per base pair. All mutations were different among the clones and none of their targets were biased to the specific bases. These findings suggest that aberrant AID expression is capable of triggering accumulation of nucleotide alterations in the p53 gene in human hepatocytes.
In this study, we provided the first evidence showing that recently identified DNA editing enzyme, AID, is significantly upregulated in the livers with several pathological settings, including chronic liver disease and HCC. In contrast, only minute AID expression is detectable in normal liver tissues. AID is a member of the cytidine deaminase family, which includes the RNA-editing enzymes, such as apolipoprotein B mRNA catalytic subunit 1 (APOBEC-1). AID was originally isolated from murine B lymphoma cells, and it has been well established that AID is specifically expressed in germinal center B cells.12 Therefore, it is reasonable to speculate that high expression of AID in HCC and the surrounding inflammatory liver tissue is derived from infiltrating B cells. However, we have clearly revealed that, in addition to infiltrating lymphocytes, human hepatocytes did express AID protein under pathological conditions.
It has been demonstrated that AID is essential for both somatic hypermutation and class switch recombination in B cells.13 Although the expression of AID is highly restricted to and regulated in germinal center B cells in healthy B cell development, one of the most striking abilities of AID was observed when it was expressed ectopically. Indeed, the aberrant expression of AID in nonlymphoid tissues leads to the accumulation of mutations, mainly at GC base pairs in nonimmunoglobulin genes.28 Furthermore, we previously reported that in addition to lymphoid malignancies, constitutive and ubiquitous expression of AID in Tg mice caused tumor development in the lung in association with high mutation frequency.19 We also found that mice overexpressing AID could develop HCCs as well as lymphoma at high frequency (unpublished data). Thus, it is tempting to speculate that the increased AID expression might be responsible for the enhanced susceptibility of the hepatocytes to somatic gene alterations, which facilitates HCC development. Our present data and hypothesis appear to be in agreement with those in previous reports showing that AID transcripts are frequently detected in various types of B-cell lymphomas.15, 16, 17, 18
An important finding in our study is the high expression levels of the AID gene in nontumor hepatocytes from livers with underlying CH and LC. Indeed, AID expression in the noncancerous liver tissues of patients with HCC was as high as that in the tumor tissue samples themselves. Livers with chronic inflammation are considered to be precursors of HCC, as more than 80% of HCCs arise in cirrhotic livers.3, 4, 5 Therefore, our data may suggest a role for ectopic AID expression in hepatocytes in hepatocarcinogenesis under chronic inflammatory conditions. The molecular mechanism of AID upregulation in the liver with chronic inflammation is unclear at present. However, it may be considered that AID expression can be induced in hepatocytes by certain inflammatory mediators produced in the inflammatory liver. Indeed, AID expression in B cells has been reported to be upregulated by TGF-β, IL-4 and CD40L.12 Since a large amount of TGF-β is present in the liver of CH and LC, and TGF-β is believed to be one of the key mediators of liver fibrosis leading to hepatocarcinogenesis,29 we examined, in this study, the effects of TGF-β on AID gene expression in hepatocytes and found that TGF-β enhanced the AID gene expression in both cultured primary human hepatocytes from normal liver tissues and human hepatoma cell lines. These findings may suggest a common mechanism for the regulation of AID gene expression in B cells and hepatocytes under inflammatory conditions, emphasizing the role of AID in hepatocarcinogenesis under inflammatory conditions.
In this study, it may be noted that the expression levels of AID in the liver specimens tended to be higher in patients with HCV infection than in those with HBV infection, although the difference was not significant. The reason for such a difference remains unknown. In this regard, Machida et al. reported that HCV infection upregulated the expression of AID transcripts in Burkitt's lymphoma and in normal peripheral blood mononuclear cells.30 Thus, it is of interest whether HCV can directly induce the upregulation of AID in hepatocytes.
Recent studies have shown that the p53 gene is a frequent target for genetic alteration in various malignancies, and the worldwide prevalence of p53 gene mutation in HCC has been reported to be around 20–52%.3, 4, 5 Notably, it has been shown that p53 gene mutations are also frequently present in nontumor tissues.31, 32, 33 For instance, multiple genetic changes in the p53 genes have been reported to be present in cytologically normal cells in premalignant conditions such as chronic inflammatory bowel disease and Barrett's esophagus.31, 32, 33 In this study, we also found that similar levels of p53 gene mutations are present in noncancerous liver and HCC tissues of the patients with HCC. This observation may be in accordance with our finding that the expression levels of AID are similar in noncancerous liver and HCC tissues, suggesting a link between the upregulation of AID and the frequent mutation of the p53 genes in the hepatocytes of the nontumorous livers of the patients with HCC. Unfortunately, however, we found no significant correlation between the expression levels of AID and the mutation frequency of the p53 gene. Similar findings were reported in human lymphoid malignancies, showing that the level of AID expression did not correlate with the presence of sequence heterogeneity in either the immunoglobulin or Bcl6 genes in diffuse large B-cell lymphomas.18 The reason for the absence of a correlation between the AID expression levels and p53 gene mutation frequencies is unknown at present. However, it may be considered that AID overexpression and the resulting somatic hypermutation may occur at the early stage of tumorigenesis.18
In this study, we found that several mutations accumulated in the p53 gene of the cells with AID activation. The mutation frequency observed in the p53 gene was relatively low when compared to those in the T-cell receptor gene reported previously.34 In addition, the mutations in the p53 gene seemed random and showed no primary sequence or motif preference in HepG2 cells with AID activation. It has been demonstrated that AID-catalyzed C deamination occurs preferentially on the sequence RGYW/WRCY [R, purine (A/G); Y, pyrimidine (C/T); W, A/T] motif in vitro.35 In this respect, it remains unclear from our study whether the p53 mutations involve the mutagenic action of aberrantly expressed AID in hepatocytes. However, we recently revealed that AID expression does not necessarily shift the base specificity of somatic hypermutation to GC and the preference of target gene by overexpression of AID is variable between the tissues or cells.34 Thus, our current findings suggest the possibility that AID is involved in the p53 mutations in human hepatocytes.
In conclusion, the present study clearly demonstrates the aberrant expression of AID in human hepatocytes with the several pathological settings. Our data suggest a role for AID upregulation in the enhancement of genetic susceptibility to mutagenesis, leading to the development of HCC in the setting of chronic liver disease.
We gratefully thank Dr. T. Honjo for his useful suggestions and critical reading of our manuscript, M. Nishikori for the generous gift of B cell lymphoma derived cells, S. Mikami for his advice in the early stages of this study and A. Kaburagi for help with the establishment of human primary hepatocyte cultures.