Introduction

Epidemiological studies have demonstrated that chronic infection with hepatitis B virus (HBV) is a major risk factor associated with development of hepatocellular carcinoma (HCC) (Beasley et al., 1981). The HBV genomic DNA is 3.2 kb in length and contains four partially overlapped open reading frames encoding S, X, C proteins, and polymerase. Viral DNA integration has been considered to play an important role in the hepatocellular carcinogenesis because the integration of HBV in host genome was observed in about 80–90% HBV-related HCC (Buendia, 1992; Robinson, 1994). Although the details of how HBV integration causes carcinogenic changes in liver are still not apparent, it is believed that HBV and host factors (e.g., inflammatory/genetic factors) may contribute together to the process of HCC development.

More than 80% cases of animal liver tumors induced by woodchuck hepatitis virus are through cis-activating myc oncogene family (Buendia, 1992; Hansen et al., 1993), yet the HBV integration site in human genome has been demonstrated as a random event (Matsubara and Tokino, 1990). The role of HBV integration played in human HCC development is more complicated and several direct or indirect mechanisms could be involved. One of the possible mechanisms is that the HBV insertion site is within or close to a tumor-related gene and may subsequently change the expression of the gene and cause HCC (Wang et al., 1990; Chami et al., 2000; Grozuacik et al., 2001). HBV integration could change the expression of viral genes and lead to overexpression of cellular oncogenes or downregulation of the expression of tumor suppressor genes (Natoli et al., 1994; Wang et al., 1995). HBV may also cause HCC by an indirect way through chronic liver injury and regeneration, which induces cirrhosis and predisposes to the development of HCC.

Although the association of HBV and HCC development has been widely studied, only a limited number of HBV integrants have been completely analysed. Therefore, additional studies with sufficient cases are needed to understand the roles of HBV integrants involved in the HCC development. In order to gain additional insight into the hepatocarcinogenesis related to HBV integration, host genomic DNA fragments containing integrated HBV sequences in 14 HCC cases were cloned and the sequences were analysed in the present study.

Results

Isolation of integrated HBV in HCC

In all, 40 HCCs and their matched nontumor liver tissues were analysed by Southern blot hybridization using HBV probe to determine the number of HBV integration and the size of host genomic DNA fragment containing the integrated HBV. HBV integration was detected in 36/40 HCCs and only in 3/40 nontumor samples. In all, 14 HCCs with HBV-integrated genomic DNA fragment less than 12 kb were selected for further study. In these 14 HCC cases, single HBV integration site was detected in eight cases. Two and three HBV integration sites were observed in four and two cases, respectively. Host genomic DNA fragments with integrated HBV sequences were isolated from all 14 constructed DNA libraries and analysed by sequencing.

HBV insertion site

All host genomic sequences flanking the 14 integrated HBV sequences were studied by sequencing and their chromosomal origins were determined in 13/14 cases using BLAST search (Figure 1 and Table 1). The chromosomal origin of flanking sequences in case H51 could not be determined because both sides contained full of human repetitive sequences. In four cases (H28, H31, H58, and H59), chromosomal location of one flanking side of HBV integrant was determined. The chromosomal location of the other side could not be determined because the cloned flanking sequence contained full of repetitive sequences. In the remaining nine cases, the chromosomal location of both flanking sides of HBV integrant was determined (Figure 1 and Table 1). The distribution of HBV integration is random; however, 9/13 insertion sites of HBV integrants were found within or close to (less than 50 bp) human repetitive sequences in this study. The detected human repetitive sequences included MER2, LINE1, Alu, LTR SSR, and alpha-satellite (Table 1).

Figure 1.

Summary of rearrangements of integrated HBV and its flanking human genome sequences in 14 HCC cases. Horizontal box represents each of the integrated HBV sequence and numbers indicate HBV nucleotide positions. Integration sites are marked in bold. Inverted triangle: interstitial deletion within the HBV sequence; shaded box: inverted and duplicated sequences; inverted sequence without duplication is marked with an asterisk (^*). Flanking solid bars represent host genomic DNA and their chromosomal locations are indicated. ND: not determined

Full figure and legend (126K)

Table 1 - Host genomic sequences flanking the HBV integrants.

Full table

Rearrangement of host genomic DNA

In the present study, the rearrangement of host genomic DNA caused by HBV integration was detected in all nine cases with known flanking DNA sequences. Based on the orientations of flanking genomic sequences of the HBV integrant, host genomic DNA change was divided into two categories: deletion and rearrangement. Deletion was determined when orientations of two flanking genomic sequences were in the same direction. Deletions of host genomic DNA in HBV insertion sites were observed in H5 (5 bp deleted) and H41 (19 bp deleted). In three cases, the flanking sites of HBV integrants were far apart (2.6 Mb for H17, 84 kb for H37, and 218 kb for H38) and the orientation of flanking sequences was opposite, implying that rearrangement but not deletion was involved. In the remaining four cases (H16, H21, H42, and H44), flanking sequences were mapped to different chromosomal regions (Figure 1 and Table 1).

Cellular genes affected by HBV integration

Cellular genes near to (within a region from 50 kb upstream to 100 kb downstream of the HBV integrant) or interrupted by HBV insertion place were also studied by BLAST search. The results are summarized in Table 2. Cellular genes interrupted by or near to integrated HBV were observed in 6/13 (46.2%) and 1/13 (7.8%) cases, respectively. Functions of most of these genes were unknown except GTF2E2 in H21 (Table 1). In the remaining six cases, no known gene was found within the region from 50 kb upstream to 100 kb downstream of the HBV integrant.

Table 2 - Cellular genes affected by HBV integration.

Full table

Rearrangements of integrated HBV sequence

Sequencing analysis showed that the lengths of integrated HBV sequences varied from each in the range 301–3527 bp. Rearrangements of integrated HBV sequences, including deletion, inversion, inverted-duplication, and duplication, were observed in all 14 cases. These results are illustrated in Figure 1 and summarized in Table 3. There were two types of deletions, deletion caused by HBV integration (deleted sequence between two insertion sites) and interstitial deletion. The deletion caused by HBV integration and interstitial deletion was detected in 13/14 and 7/14 HBV integrants, respectively. Complex rearrangements were detected in five HBV integrants. In H28, an inverted-duplication (nt 2390–2858) was found besides the deletion (nt 1809–2307) caused by HBV integration. In H38, a duplication (nt 1610–1715) was observed besides an interstitial deletion (nt 2322–2521). In H42, a deletion caused by integration (nt 1696–1728), an interstitial deletion (nt 1794–1817), an inversion (nt 1729–1793), and a duplication (nt 2954–3215/1–107) were detected. In H58, a deletion caused by HBV integration (nt 1571–1791), two interstitial deletions (nt 2992–3116; nt 3190–3187), and an inverted-duplication (nt 2722–2990) were observed. In H59, a deletion caused by HBV integration (nt 2508–3215/1–41), two interstitial deletions (nt 1450–1480; nt 2139–2265), and an inverted-duplication (nt 2161–2439) were observed (Figure 1).

Table 3 - Summary of HBV sequence rearrangements in 14 cases of HCC.

Full table

Rearrangements of four HBV genes are also summarized in Table 3. The most frequent rearrangement was detected in X gene (in 12/14 HBV integrants). Rearrangement of S gene (including pre-S and S) was observed in 7/14 HBV integrants. Rearrangements involving C and P genes were observed in 7/14 and 8/14 of HBV integrants, respectively.

Rearrangements of integrated X gene

In the present study, X gene rearrangements were detected in 12/14 cases. Among these, 10/12 rearrangements resulted in 3'-deleted X gene. X protein is composed of 154 amino acids. In seven cases (H5, H17, H28, H31, H38, H41, and H42), the deleted part of X protein was less than 48 amino acids in the C-terminal. In H21, H58, and H59, the deleted part of X protein was 65, 90, and 129 amino acids, respectively (Figure 1 and Table 3).

HBV X gene in nontumor liver tissue

X gene of free replicated HBV in matched nontumor liver tissue in all 14 cases was also amplified by PCR using primers X1 and CR2 and no rearrangement was found. Figure 2 shows some examples of PCR results. All X genes in nontumor liver tissues showed no rearrangement; however, deletion and insertion of X gene were detected in two and one HCC cases, respectively. In cases H31 and H51, no PCR products of X gene were observed in HCC because of the deletion of X gene. In H37, a normal size of PCR product was detected in both HCC and its matched nontumor counterpart. Sequencing result also confirmed that no rearrangement of X gene was found in HCC. In H59, a larger sized PCR product was observed in HCC and our sequencing study demonstrated that a 278 bp insertion (nt 2161–2439) was found in X gene. X gene in nontumor liver tissues of all 14 cases was also studied by sequencing and no rearrangement was observed (data not shown).

Figure 2.

Examples of rearrangements of X gene in HCC caused by HBV integration. X gene in four pairs of tumors (T) and their matched nontumor counterparts (N) were PCR amplified and the expected product is 664 bp. Normal size PCR product is observed in all nontumor samples and HCC sample in H37. PCR product is absent in HCC sample in H31 and H51 because of the deletion of the X gene. In H59, a larger PCR product is detected in HCC sample caused by a 278 bp insertion

Full figure and legend (50K)

Discussion

Although the HBV integration has been associated with HCC development, the molecular mechanism of HBV integration in the pathogenesis of HCC remains unclear. In the present study, no common chromosomal region was involved in HBV integration in 14 HCC cases. This result is consistent with the previous reports (Tokino and Matsubara, 1991; Grozuacik et al., 2001), implying that HBV integration is a random event. However, the insertion site of HBV in 10/14 cases was found either within or close to a human repetitive sequence. This observation suggests that HCC integration might be sequence dependent, although no more data exist to support this point in this study.

Cellular gene interrupted by or near to HBV integrant was observed in 7/13 cases. Most of these genes (6/7) are functional unknown genes. The only known gene is GTF2E2, encoding the p34 small subunit of the general transcription factor TFIIE (Peterson et al., 1991). TFIIE regulates TFIIH activity, which is involved in the eucaryotic nucleotide excision–repair pathway (Jaspers and Hoeijmakers, 1995). The association of GTF2E2 and carcinogenesis is unclear. Our observation in this study showed that it is unlikely that the HBV integration is able to affect directly the host cancer-related gene and lead to the development of HCC.

Another interesting finding is that in 4/9 cases (44.4%) with known junction sequences at both sites, flanking host genomic DNA sequences were mapped to different chromosomes. This phenomenon has also been reported previously (Ziemer et al., 1985; Su et al., 1998; Brechot et al., 2000). It is likely that the integrated HBV may possess the capacity to be excised from one chromosome and then transposed to another chromosome region through reintegration. Integration of HBV can cause deletion or rearrangement of host genomic DNA during its insertion into the host genome. This kind of deletion was observed in all five cases with both junction sequences from the same chromosome. The deleted or rearranged region ranged from several base pairs to 2.6 Mb, implying that the integration of HBV into host genome may induce chromosome instability, which has also been suggested by others (Wang and Rogler, 1988; Urano et al., 1991).

The most interesting finding in this study is that rearranged X protein caused by HBV integration was observed in 12/14 cases. Among these cases, 3'-deleted X gene was detected in 10 cases. X protein (HBx) has been considered to be strongly implicated in hepatocarcinogenesis, although its precise role remains uncertain. Modulatory function of HBx has been widely studied and been associated with transcription, signaling pathways, cell-cycle control, apoptosis, and carcinogenesis (for a review see: Arbuthnot et al., 2000; Yeh, 2000; Diao et al., 2001; Murakami, 2001). Biological evidence for the oncogenic function of HBx is derived from the fact that it can induce HCC in vivo in certain lines of transgenic mice (Kim et al., 1991; Yu et al., 1999). However, the direct oncogenic role of HBx was not supported by several other transgenic studies (Billet et al., 1995; Reifenberg et al., 1997; Terradillos et al., 1998; Yu et al., 1999; Madden et al., 2001). Moreover, some studies suggest that the wild-type HBx may act as a tumor suppressor gene. Sirma et al. (1999) reported that HBx could inhibit clonal outgrowth of cells and induces apoptosis by a p53-independent pathway. Furthermore, they found that HBx could induce a late G1 cell cycle block. Recently, Tu et al. (2001) found that full-length HBx could suppress the focus formation induced by the cooperation of ras and myc oncogenes, while COOH-terminally truncated HBx can enhance the transforming ability of ras and myc.

The oncogenic ability of truncated X protein might be associated with the losses of p53-dependent transcriptional repression binding site at 1637–1667 nt in the enhancer II region (Lee et al., 1998) and SP1 binding sites at 1733–1754 nt in the core promoter region (Li and Ou, 2001). Previous study has shown that p53 negatively regulates HBV replication and inhibits the expression of all four HBV genes. p53-dependent transcriptional repression binding site was then localized in HBV enhancer II region (Lee et al., 1998). The expression of HBV genes is regulated by a number of transcription factors including Sp1, and the HBV core promoter region contains two Sp1 binding sites (SP1-1 and SP1-2, Figure 3). SP1-2 site was found to regulate negatively the transcription of S and X genes and so the expression of these two genes was increased when SP1-2 was removed (Li and Ou, 2001). Another recent study showed that COOH-terminal region of X protein had inhibitory effects on cell proliferation and transformation, and deletion of this region may enhance the transformation ability of ras and myc (Tu et al., 2001). Deletion of 14 aa in COOH-terminal was enough to abrogate the suppressive effects of HBx on cell growth.

Figure 3.

Distribution of the deletion breakpoints within X gene (nt 1376–1840) in 10 HBV integrants. Breakpoints are indicated by arrows. p53-dependent transcriptional repression binding site, transcription factor Sp1 binding sites, and cell growth-suppressive effect domain are underlined

Full figure and legend (78K)

In the present study, 3'-deleted X gene was observed in 10/14 integrants and the distributions of the deletion breakpoints are summarized in Figure 3. Among these deletions, losses of p53-dependent transcriptional repression binding site and Sp1 binding site were detected in two and four cases, respectively. Loss of growth-suppressive effect domain was observed in all integrants and the minimum deleted region is 25 nt (9 aa). To summarize, our results strongly suggest that the 3'-deleted X gene and consequent C-terminal truncation of HBx caused by the HBV integration plays an important role in the HCC development. Further investigation may reveal the molecular mechanism of C-terminal truncated HBx in the development of HCC.

Materials and methods

Patients and tumor samples

In total, 40 HCC tumors with serum HBsAg-positive and their matched surrounding nontumor liver tissues were obtained from the hepatosectomy at Eastern Hepatobiliary Surgery Hospital, Shanghai, China. Among these 40 HCC patients, 33 were males and seven were females. The age range of these cases was from 17 to 71 years, with a mean age of 47 years. The tumor size ranged from 3.5 to 15 cm, with a mean size 7.7 cm. Chronic cirrhoses were detected in 34/40 cases (85%). Only 14/40 cases were selected for further study and their clinical information is summarized in Table 4.

Table 4 - Clinical, histological, and serological profiles of 14 HCC cases.

Full table

Southern blot hybridization

Genomic DNA was extracted from frozen samples by proteinase K/sodium dodecyl sulfate digestion followed by phenol/chloroform/isoamyl and alcohol extraction. Southern blot hybridization was performed by standard method. The probe used for detecting the HBV integration was 1.8 kb HBV genomic DNA containing X and C genes. A measure of 10 g of genomic DNA was digested with EcoRI, fractionated on a 1% agarose gel, transferred to a nylon membrane (Bio-Rad, Hercules, CA, USA), and hybridized overnight at 42°C with ³²P-labeled probes.

Construction of genomic DNA libraries and library screening

After the size of the EcoRI-digested genomic DNA fragment containing integrated HBV was determined by Southern blot analysis, genomic DNA library was constructed to isolate the integrated HBV fragment. A measure of 5 g of host genomic DNA was completely digested with EcoRI, purified with microcom (Millipore Corporation, Bedford, MA, USA), and cloned into lambda phage vector ZAPII (Stratagene, La Jolla, CA, USA). The cloned genomic DNA was packaged in vitro using Gigapack Gold extract (Stratagene) and then transformed into E. coli XL1-Blue MRF'.

The same HBV probe used for Southern blot hybridization was used to screen the constructed libraries (1–4.8 10⁶ PFU) for isolating HBV integrants. The probe was labeled with ³²P and hybridized to the library plaque lifts by the standard method.

Sequencing

The isolated genomic DNA fragment containing HBV integrant was analysed by direct sequencing using the ABI310 automatic sequencer (Applied Biosystems, Foster City, CA, USA). A total of 13 pairs of primers (Table 5) for sequencing integrated HBV DNA were used in the analysis. HBV primers were designed based on sequence from HBV clone HBV-ASA-EX3 (Accession # D50519.1). Raw sequencing data were imported using EditSeq V4.05 (DNASTAR Inc., Madison, WI, USA) and sequence assembly was carried out using SeqMan II (DNASTAR Inc.).

Table 5 - Sequencing primers for HBV integrants.

Full table

NCBI BLAST and Map Viewer search

Sequences derived from host flanking sites of HBV integrants were used to search for the corresponding bacterial artificial chromosome (BAC) from NCBI BLAST database. Mapping of the BAC sequence was fulfilled by searching NCBI Map Viewer database.

Amplification of HBV X gene

In order to determine the differences between the X gene of the integrated HBV in HCC and that of free replicated HBV in nontumor liver tissue, X gene from nontumor liver tissue was amplified by PCR (forward primer (X1): 5'-CAG CTTGTTTTGCTCGCAGC (nt 1286–1305), reverse primer (CR2): 5'-GAGTAACTCCACAGAAGCTC (nt 1929–1948)) and then analysed by agarose gel electrophoresis and sequencing.

References

1. Arbuthnot P, Capovilla A and Kew M. (2000). J. Gastroenterol. Hepatol., 15, 357–368. | Article | PubMed | ChemPort |
2. Beasley RP, Hwang LY, Lin CC and Chien CS. (1981). Lancet, 2, 1129–1133. | Article | PubMed | ISI | ChemPort |
3. Billet O, Grimber G, Levrero M, Seye KA, Briand P and Joulin V. (1995). J. Virol., 69, 5912–5916. | PubMed |
4. Brechot C, Gozuacik D, Murakami Y and Brechot PP. (2000). Semin. Cancer Biol., 10, 211–231. | Article | PubMed | ISI | ChemPort |
5. Buendia MA. (1992). Adv. Cancer Res., 59, 167–226. | PubMed |
6. Chami M, Gozuacik D, Saigo K, Capiod T, Falson P, Lecoeur H, Urashima T, Beckmann J, Gougeon ML, Claret M, Maire M, Brechot C and Paterlini-Brechot P. (2000). Oncogene, 19, 2877–2886. | Article | PubMed | ChemPort |
7. Diao J, Garces R and Richardson CD. (2001). Cytokine Growth Factor Rev., 12, 189–205. | Article | PubMed | ISI | ChemPort |
8. Grozuacik D, Murakami Y, Saigo K, Chami M, Mugnier C, Lagorec D, Okanoue T, Urashima T, Brechot C and Paterlini-Brechot P. (2001). Oncogene, 20, 6233–6240. | Article | PubMed | ChemPort |
9. Hansen LJ, Tennant BC, Seeger C and Ganem D. (1993). Mol. Cell. Biol., 13, 659–635. | PubMed |
10. Jaspers NGJ and Hoeijmakers JHJ. (1995). Curr. Biol., 5, 700–702. | PubMed |
11. Kim CM, Koike K, Saito I, Miyamura T and Jay G. (1991). Nature, 351, 317–320. | Article | PubMed | ISI | ChemPort |
12. Lee H, Kim HT and Yun Y. (1998). J. Biol. Chem., 273, 19786–19791. | PubMed |
13. Li J and Ou J-H. (2001). J. Virol., 75, 8400–8406. | Article | PubMed | ISI | ChemPort |
14. Madden CR, Finegold MJ and Slagle BL. (2001). J. Virol., 75, 3851–3858. | Article | PubMed | ISI | ChemPort |
15. Matsubara K and Tokino T. (1990). Mol. Biol. Med., 7, 243–260. | PubMed |
16. Murakami S. (2001). J. Gastroenterol., 36, 651–660. | Article | PubMed | ISI | ChemPort |
17. Natoli G, Avantaggiati ML, Chirillo P, Puri PL, Ianni A, Balsano C and Levrero M. (1994). Oncogene, 9, 2837–2843. | PubMed | ISI | ChemPort |
18. Peterson MC, Inostroza J, Maxon ME, Flores O, Admon A, Reinberg D and Tjian R. (1991). Nature, 354, 369–373. | Article | PubMed | ISI | ChemPort |
19. Reifenberg K, Lohler J, Pudollek HP, Schmitteckert E, Spindler G, Kock J and Schlicht HJ. (1997). J. Hepatol., 26, 119–130. | PubMed |
20. Robinson WS. (1994). Ann. Rev. Med., 45, 297–323. | PubMed |
21. Sirma H, Giannini C, Poussin K, Paterlini P, Kremsdorf D and Brechot C. (1999). Oncogene, 18, 4848–4859. | Article | PubMed | ISI | ChemPort |
22. Su TS, Hwang WL and Yauk YK. (1998). DNA Cell Biol., 17, 415–425. | PubMed |
23. Terradillos O, Pollicino T, Lecoeur H, Tripodi M, Gougeon ML, Tiollais P and Buendia MA. (1998). Oncogene, 17, 2115–2123. | Article | PubMed | ISI | ChemPort |
24. Tokino T and Matsubara K. (1991). J. Virol., 65, 6761–6764. | PubMed | ChemPort |
25. Tu H, Bonura C, Giannini C, Mouly H, Soussan P, Kew M, Paterlini-Brechot P, Brechot C and Kremsdorf D. (2001). Cancer Res., 61, 7803–7810. | PubMed | ChemPort |
26. Urano Y, Watanabe K, Lin C, Hino O and Tamaoki T. (1991). Cancer Res., 51, 1460–1464. | PubMed |
27. Wang HP and Rogler CE. (1988). Science, 48, 72.
28. Wang J, Chenivesse X, Henglein B and Brechot C. (1990). Nature, 343, 555–557. | Article | PubMed | ISI | ChemPort |
29. Wang XW, Gibson MK, Vermeulen W, Yeh H, Forrester K, Sturzbecher HW, Hoeijmakers JH and Harris CC. (1995). Cancer Res., 55, 6012–6016. | PubMed | ISI | ChemPort |
30. Yeh CT. (2000). J. Gastroenterol. Hepatol., 15, 339–341. | PubMed |
31. Yu DY, Moon HB, Son JK, Jeong S, Yu SL, Yoon H, Han YM, Lee CS, Park JS, Lee CH and Hyun BH. (1999). J. Hepatol., 31, 123–132. | Article | PubMed | ChemPort |
32. Ziemer M, Garcia P, Shaul Y and Rutter WJ. (1985). J. Virol., 53, 885–892. | PubMed | ISI | ChemPort |

Acknowledgements

This study is supported in part by the Leung Kwok Tze Foundation and National Natural Science Foundation of China (Grant number: 30171046).