Potential conflict of interest: Nothing to report.
This work was funded by the National Natural Science Foundation of China (No. 30801328, 81472770), Science Fund for Creative Research Groups (No. 81221061, 81201940), Shanghai Pujiang Project, and New Excellent Youth Plan (XYQ2013074) for financial support.
Hepatocellular carcinoma (HCC) is a common solid tumor and the third leading cause of cancer death worldwide.1 Hepatitis B virus (HBV) is a major etiological agent in China, Southeast Asia, and sub‐Saharan Africa, and individuals with chronic HBV infection are at increased risk of developing HCC, particularly those with chronic liver disease and cirrhosis.2 Average age at onset of HBV‐associated HCC is 50 years3; thus, the recommendations advise HCC screening for Asian male HBV patients older than 40 and Asian female HBV patients older than 50.5 Nonetheless, incidence of HCC in patients younger than 40, especially in high‐risk populations, is relatively high.6 Recent studies have reported a significant prevalence and worse prognosis in early‐onset HCC patients,8 suggesting that there may be different mechanisms of hepatocarcinogenesis between early‐ and late‐onset HCC.10 However, there are limited studies focusing on the etiology of early‐onset HCC.
Given that the hepatocarcinogenic process involves interplay between HBV and host hepatocytes, both genomes may contribute to the final pathogenic outcomes, either individually or synergistically. With regard to the hepatitis B viral factors, the genotype has been increasingly recognized as affecting the clinical outcome of chronic HBV infection, including cirrhosis and HCC.11 Some data have associated genotype B HBV with the development of HCC in young carriers in Taiwan,13 and Livingston et al.14 found that median age of HCC patients with genotype F was much lower than that of patients with other genotypes, indicating that the HBV subgenotype might contribute to early onset of HCC.
In addition, a major proposed pathway by which HCC may arise from chronic HBV infection is integration of HBV DNA into the host genome, resulting in oncogene activation, tumor‐suppressor gene inactivation, or other predisposition to chromosomal instability.15 Many studies have detected evidence of HBV DNA integrated specifically into HCC cells, supporting the hypothesized role of HBV‐DNA integration in hepatocarcinogenesis. However, although conventional polymerase chain reaction (PCR)‐based methods, such as Alu‐HBV PCR, can be used to detect the presence of viral integration, these methods are significantly biased toward Alu regions, and only a small subset of insertions can be detected or only insertions close to viral sequences can be found.16
To investigate the mechanisms underlying early‐onset HCC, we characterized the integration patterns and genotypes of HBV in early‐ and late‐onset HCC patients. We used a high‐throughput viral integration detection (HIVID) method that can effectively detect HBV integration at a single base‐pair (bp) resolution.19 We found that HBV B2 is predominantly present in early‐onset HCC, whereas HBV C2 is mainly found in late‐onset HCC. HBV integration was also found to be a common phenomenon, even in early‐onset HCC, but the hotspots between early‐ and late‐onset disease are rather different. A breakpoint between c‐Myc and PVT1, located in the 8q24 gene desert, was frequently detected in early‐onset HCC, resulting in overexpression of c‐MYC, plasmocytoma variant translocation 1 (PVT1), and microRNA (miR)−1204 in tumors that might contribute to the development of early‐onset HCC.
Materials and Methods
HCC Patient Samples
For HBV genotype analysis, we thoroughly studied 258 Chinese individuals diagnosed with HCC who underwent curative primary hepatectomy at Eastern Hepatobiliary Surgery Hospital (Shanghai, China). Demographic and clinicopathological data for these individuals are summarized in Table 1. These patients were divided into two groups: 113 early‐onset HCC (diagnosis at age <30) and 145 late‐onset HCC (diagnosis at age ≥70). All patients had been diagnosed with HCC with concurrent HBV infection and without for autoimmune hepatitis and metabolic and/or genetic disorders, such as Wilson's disease, hemochromatosis, and primary biliary cirrhosis. No patient had a history of alcoholism, aflatoxin exposure, or hepatitis C virus (HCV) coinfection. Patients with liver cirrhosis (LC) were diagnosed on the basis of histological findings as diffuse fibrosis and formation of pseudolobuli, as well as ultrasonic features of LC plus evidence of hyperplenism. The study protocol conformed to the 1975 Delcaration of Helsinki and was approved by the ethics committees of the Second Military Medical University (Shanghai, China), and an informed written consent was obtained from each patient.
Table 1 -
Clinicopathological Features of Patients With HCC for Genotype and Integration Analysis
|
HBV Genotype Analysis |
HBV Integration Analysis |
| Early‐Onset (n = 113) |
Late‐Onset (n = 145) |
P Value |
Early‐Onset (n = 30) |
Late‐Onset (n = 30) |
P Value |
| Age, years |
22.8 ± 6.5 |
72.8 ± 8.0 |
|
20.6 ± 4.2 |
74.2 ± 3.0 |
|
| Gender (%) |
|
|
NS |
|
|
NS |
| Male |
102 (90.3) |
128 (88.3) |
|
24 (80) |
26 (86.7) |
|
| Female |
11 (9.7) |
17 (11.7) |
<0.01
|
6 (20) |
4 (13.3) |
<0.01
|
| Genotype (%) |
|
|
|
|
|
|
| A |
3 (2.7) |
8 (5.5) |
|
0 |
0 |
|
| B2 |
54 (47.8) |
38 (26.2) |
NS |
22 (73.3) |
6 (20) |
NS |
| C1 |
18 (15.9) |
12 (8.3) |
NS |
8 (26.7) |
24 (80) |
NS |
| C2 |
38 (33.6) |
87 (60.0) |
|
0 |
0 |
|
| Tumor size, cm |
8.8 ± 2.3 |
8.7 ± 3.0 |
NS |
8.4 ± 2.8 |
8.6 ± 1.2 |
NS |
| Tumor number (%) |
|
|
|
|
|
|
| Single |
82 (72.6) |
108 (74.4) |
NS |
22 (73.3) |
21 (70.0) |
NS |
| Multiple |
31 (27.4) |
37 (25.6) |
NS |
8 (26.7) |
9 (30) |
NS |
| Alpha‐fetoprotein, ng/mL (%) |
|
|
NS |
|
|
NS |
| ≥400 |
56 (49.5) |
78 (53.8) |
NS |
14 (46.7) |
15 (50.0) |
NS |
| <400 |
57 (50.5) |
67 (46.2) |
NS |
16 (53.3) |
15 (50.0) |
NS |
| Albumin, g/dL |
3.7 ± 4.9 |
3.9 ± 3.7 |
<0.001
|
3.6 ± 4.2 |
3.4 ± 2.8 |
<0.01
|
| Platelets, 103/uL |
183.7 ± 62.4 |
192.3 ± 79.6 |
NS |
176.4 ± 48.2 |
188.0 ± 72.4 |
NS |
| Alanine aminotransferase, u/L |
58.5 ± 55 |
62.4 ± 49 |
|
60.2 ± 45 |
64.3 ± 38 |
|
| Total bilirubin, umol/L |
14.8 ± 4.8 |
15.1 ± 7.8 |
NS |
14.6 ± 3.1 |
14.8 ± 6.2 |
NS |
| Prothrombin time, seconds |
13.2 ± 2.3 |
13.4 ± 1.4 |
NS |
13.3 ± 1.4 |
13.6 ± 1.8 |
NS |
| Cirrhosis (%) |
72 (63.7) |
121 (83.4) |
|
19 (63.6) |
24 (80.0) |
|
| Child‐Pugh class (%) |
|
|
NS |
|
|
NS |
| A |
103 (91.2) |
139 (96) |
NS |
26 (86.7) |
27 (90.0) |
NS |
| B |
10 (8.8) |
6 (4.0) |
<0.001
|
4 (12.3) |
3 (10.0) |
<0.001
|
| HBeAg positivity |
65 (57.5) |
77 (53.1) |
NS |
16 (53.3) |
14 (46.7) |
NS |
| Preoperative DNA level, IU/mL (%) |
|
|
|
|
|
|
| ≤1,000 |
69 (61.1) |
86 (59.3) |
|
17 (56.7) |
16 (53.3) |
|
| >1,000 |
44 (38.9) |
59 (40.7) |
|
13 (43.3) |
14 (46.7) |
|
| Presence of satellite nodules (%) |
16 (14.2) |
23 (15.9) |
|
3 (10.0) |
4 (13.3) |
|
| Presence of vascular invasion |
12 (10.6) |
11 (7.6 %) |
|
5 (16.7) |
4 (13.3) |
|
| Median survival time (months) |
22.0 ± 9.8 |
44.1 ± 14.2 |
|
19.0 ± 10.4 |
40.3 ± 13.8 |
|
| BCLC stage (%) |
|
|
|
|
|
|
| 0 |
12 (10.6) |
18 (12.4) |
|
4 (13.3) |
3 (10) |
|
| A |
60 (53.1) |
66 (45.5) |
|
16 (53.3) |
14 (46.7) |
|
| B |
31 (27.4) |
44 (30.3) |
|
8 (26.6) |
10 (33.3) |
|
| C |
10 (9.7) |
17 (11.7) |
|
2 (6.6) |
3 (10.0) |
|
Abbreviations: HBeAg, hepatitis B envelope antigen; BCLC, the Barcelona‐Clínic Liver Cancer (BCLC) staging system; NS, not significant. P value <0.05 is shown in bold.
We randomized selected 30 patients in the early‐onset group and 30 patients in the late‐onset group for HBV capture sequencing analysis. Demographic and clinicopathological data for these individuals are also summarized in Table 1.
HBV Capture Sequencing
We used HIVID to detect HBV integration.19 Experimental procedures are described in Fig. 1 and Supporting Materials and Methods.
;)
Figure 1: Experimental process of HBV virus capture sequencing.
Figure 2: Distribution of genotypes of HBV in early‐ and late‐onset patients.
PCR and Sanger Sequencing Validation
PCR primers giving products of approximately 200 bp were designed to capture the breakpoints of each HBV integration site. All PCR experiments were performed with tumors on a GeneAmp PCR System 9700 thermal cycler (Life Technologies, Grand Island, NY). For each PCR product, we used Applied Biosystems (Foster City, CA) 3730 DNA analyzers to perform dye terminator Sanger sequencing.
Genotyping of HBV
Determination of HBV genotypes and subgenotypes HBV DNA was extracted from 200 uL of serum of HCC patients using High Pure Viral Nucleic Acid kits (Roche Diagnostics GmbH, Roche Applied Science, Penzberg, Germany) according to the manufacturer's instruction. Tumor samples were also used as the sources of HBV DNA if insufficient serum was obtained. HBV genotypes and subgenotypes were determined by using a multiplex PCR assay.12 HBV genotypes of samples with low levels of HBV DNA were identified by nested multiplex PCR.
RNA Isolation and Quantitative Real‐Time Reverse‐Transcriptase PCR
RNA from tumor (T) and paired adjacent nontumor (NT) samples were isolated as described in a previous work.21 Real‐time PCR was performed using complementary DNA as a template and a standard SYBR‐Green PCR kit on a StepOne Plus system (detailed in Supporting Materials and Methods).
Statistical Analysis
Data are presented as means ± standard deviation for continuous data frequency (n) and percentage for categorical data. To compare the values between the two groups, chi‐square or Fisher's exact tests were performed for categorical variables, as appropriate. Mann‐Whitney's U test was used for continuous variables. All statistical tests were two sided. P < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 18 for Windows (SPSS, Inc., Chicago, IL).
Results
Subgenotype B2 Is Predominately Found in Early‐Onset Patients
We surveyed HBV genotypes using DNA extracted from serum or tumor tissues of 113 early‐onset (≤30) and 145 late‐onset (≥70) patients with HBV‐associated HCC who underwent curative primary hepatectomy in Eastern Hepatobiliary Surgery Hospital. Demographic and clinicopathological data for these individuals are summarized in Table 1. Compared with late‐onset patients, early‐onset HCC patients display a reduced incidence of cirrhosis (63.7% vs. 83.4%; P < 0.001). In addition, although no significant differences in the male‐to‐female ratios, tumor size and number, and other clinicopathological features were observed in the early‐ and late‐onset groups, median survival time for early‐onset patients was significantly reduced, compared with late‐onset patients (P < 0.001; Table 1).
Of the 258 patients with HBV‐associated HCC, we found 11 (4.3%) patients infected with HBV genotype A, 92 (35.7%) with subgenotype B2, 30 (3.9%) with subgenotype C1, and 125 (48.4%) with subgenotype C2. These results indicate that subgenotypes B2 and C2 are the predominant subgenotypes in East China. Intriguingly, compared with the late‐onset patients, subgenotype B2 was found in 47.8% (54 of 113) of the early‐onset patients, though in only 26.2% (38 of 145) of the late‐onset group. In contrast, subgenotype C2 was found in 33.6% (38 of 113) of the early‐onset patients and was significantly increased to 60% (87 of 145) in the late‐onset patients. These data suggested that HBV subgenotype distribution in early‐ and late‐onset HCC was significantly different (P < 0.01), consistent with data from patients in Taiwan12 and Eastern China.22
HBV Integration Commonly Occurs in Both Early‐ and Late‐Onset HCC
Case‐control studies demonstrate that the HBV genotype C is associated with an increased risk of HCC, compared with the genotype B in Japan23 and Taiwan.12 However, median survival time suggests a poorer early‐onset HCC prognosis, compared with the late‐onset group. These seemingly contradictory results imply that the mechanisms underlying early‐ and late‐onset HCC development might differ.
To investigate whether HBV integration contributes to early‐onset HCC, we randomly selected 30 patients in the early‐onset group and 30 in the late‐onset group and detected HBV integration events in early‐ and late‐onset HCC tumors using an unbiased HIVID. For comparison and analysis purposes, we ensured that the major clinicopathological features, such as ratio of LC, between individuals for HIVID and individuals for HBV genotype analysis are consistent (Table 1).
We first extracted tumor genomic DNA (gDNA) from 30 early‐onset and 30 late‐onset HCCs. Each group was divided into six pooled samples containing an equal amount of gDNA from 5 HCC tumors. Then, we detected HBV integration breakpoints in pooled samples by HIVID. We generated ∼1.8 Gb of sequence data for each pooled sample. Approximately 9 million paired‐end reads were obtained in the analysis. Approximately 85.6% and 0.09% of these reads were aligned to the human genome and HBV genome, respectively. The ratio of the reads mapped to human genome to the HBV genome is 951 (85.6%/0.09%). Given that the human genome is approximately 106‐fold greater than the HBV genome, HBV sequences are significantly increased after HBV capturing. Finally, after paired‐end assembly and remapping, we obtained an average of 1,928 reads supporting HBV integration breakpoints for each pooled sample.
A total of 97 HBV integration breakpoints were detected in six pooled early‐onset samples and 104 integration events in six late‐onset samples by HIVID, with each supported by at least two paired‐end reads. Next, we detected the integration sites obtained from the pooled samples in each separate sample using PCR and sequencing. We successfully validated 74.3% (72 of 97) breakpoints in 23 early‐onset HCC patients and 81.6% (84 of 104) breakpoints in 27 late‐onset patients as real integration sites, indicating that this approach has a high sensitivity and specificity for the detection of HBV integration.
HBV integration events were detected in 76.7% (23 of 30) of early‐onset HCC patients, and the average number of integration breakpoints in early‐onset HCC is 3.13 (range, 1‐6). In late‐onset HCC, the integration detection rate is 90% (27 of 30), and the average number of integration breakpoints is 3.11 (range, 1‐5). Collectively, these results suggest that HBV integration commonly occurs in both early‐ and late‐onset HCC (Supporting Tables 1 and 2).
Table 2 -
Recurrent Hotspots of HBV Integration in Late‐ and Early‐Onset Groups
| Gene |
Human_Position |
Function |
Virus_Position |
Affected HBV Gene |
Affected Patients |
| Late‐onset group |
| STAT1 |
chr2: 191836387 |
Intronic |
3215 |
Polymerase; pre‐S1/pre‐S2/S |
G‐17T |
|
chr2: 191836431 |
Intronic |
2936 |
Polymerase; pre‐S1/pre‐S2/S |
G‐21T |
|
chr2: 191836431 |
Intronic |
1403 |
Polymerase; X protein |
G‐11T |
|
chr2: 191836589 |
Intronic |
1412 |
Polymerase; X protein |
G‐23T |
| ALB |
chr4: 74271597 |
Intronic |
24 |
Polymerase; pre‐S1/pre‐S2/S |
G‐3T |
|
chr4: 74271706 |
Intronic |
2936 |
Polymerase; pre‐S2/S |
G‐7T |
|
chr4: 74271597 |
Intronic |
24 |
Polymerase; pre‐S1/pre‐S2/S |
G‐9T |
|
chr4: 74271706 |
Intronic |
2936 |
Polymerase; pre‐S2/S |
G‐27T |
| TERT |
chr5: 1295378 |
Promoter |
1560 |
Polymerase; X protein |
G‐2T |
|
chr5: 1295468 |
Promoter |
2318 |
Polymerase; pre‐core/core |
G‐5T |
|
chr5: 1295741 |
Promoter |
1831 |
X protein; pre‐core/core |
G‐1T |
|
chr5: 1295737 |
Promoter |
1817 |
X protein; pre‐core/core |
G‐13T |
|
chr5: 1295549 |
Promoter |
911 |
Polymerase |
G‐16T |
|
chr5: 1295500 |
Promoter |
460 |
Polymerase; S |
G‐18T |
|
chr5: 1295541 |
Promoter |
1442 |
Polymerase; X protein |
G‐14T |
|
chr5: 1295534 |
Promoter |
1288 |
Polymerase |
G‐22T |
|
chr5: 1295529 |
Promoter |
1136 |
Polymerase |
G‐19T |
|
chr5: 1295516 |
Promoter |
400 |
Polymerase; S |
G‐24T |
|
chr5: 1295379 |
Promoter |
1561 |
Polymerase; X protein |
G‐23T |
| ST18 |
chr8: 53048426 |
Intronic |
1589 |
Polymerase; X protein |
G‐25T |
|
chr8: 53048618 |
Intronic |
510 |
Polymerase; S |
G‐17T |
|
chr8: 53048633 |
Intronic |
1468 |
Polymerase; S |
G‐15T |
|
chr8: 53048633 |
Intronic |
630 |
Polymerase; S |
G‐28T |
| STXBP4 |
chr17: 53048162 |
Intronic |
1461 |
Polymerase; X protein |
G‐25T |
|
chr17: 53048221 |
Intronic |
1342 |
Polymerase |
G‐29T |
| PCMTD2 |
chr20: 62918375 |
Intronic |
1818 |
pre‐core/core; X protein |
G‐6T |
|
chr20: 62918250 |
Intronic |
666 |
Polymerase; S |
G‐7T |
|
chr20: 62918250 |
Intronic |
1486 |
Polymerase; X protein |
G‐21T |
| MLL4 |
chr19: 36212641 |
Intronic |
1467 |
Polymerase; X protein |
G‐1T |
|
chr19: 36212743 |
Intronic |
400 |
Polymerase; S |
G‐10T |
|
chr19: 36212721 |
Intronic |
504 |
Polymerase; S |
G‐21T |
|
chr19: 36212743 |
Intronic |
446 |
Polymerase; S |
G‐27T |
| GPR116 |
chr6: 46816782 |
Promoter |
489 |
Polymerase; S |
G‐21T |
|
chr6: 46816782 |
Promoter |
2225 |
pre‐core/core |
G‐1T |
|
chr6: 46816770 |
Promoter |
1244 |
Polymerase |
G‐20T |
| Early‐onset group |
| FN1 |
chr2: 216248136 |
Exonic |
2010 |
pre‐core/core |
L‐12T |
|
chr2: 216248919 |
Exonic |
1932 |
pre‐core/core |
L‐14T |
|
chr2: 216248944 |
Intronic |
967 |
Polymerase |
L‐24T |
|
chr2: 216248976 |
Intronic |
2032 |
pre‐core/core |
L‐27T |
| SYT12 |
chr11: 66807124 |
Intronic |
1314 |
Polymerase |
L‐11T |
|
chr11: 66807269 |
Intronic |
1589 |
Polymerase; X protein |
L‐7T |
| MYC/PVT1 |
chr8: 128764911 |
Integenic |
660 |
Polymerase; S |
L‐15T |
|
chr8: 128764911 |
Integenic |
660 |
Polymerase; S |
L‐17T |
|
chr8: 128764911 |
Integenic |
660 |
Polymerase; S |
L‐5T |
|
chr8: 128764911 |
Integenic |
660 |
Polymerase; S |
L‐16T |
|
chr8: 128764911 |
Integenic |
660 |
Polymerase; S |
L‐22T |
|
chr8: 128764911 |
Integenic |
660 |
Polymerase; S |
L‐29T |
|
chr8: 129229133 |
Integenic |
2387 |
Polymerase |
L‐21T |
| GATM |
chr15: 45663260 |
Intronic |
1798 |
X protein |
L‐19T |
|
chr15: 45663260 |
Intronic |
757 |
Polymerase |
L‐23T |
|
chr15: 45661997 |
Intronic |
92 |
Polymerase; Pre‐S2/S, S |
L‐26T |
| ME1 |
chr6: 83965334 |
Intronic |
835 |
Polymerase; S |
L‐3T |
|
chr6: 83965334 |
Intronic |
1825 |
pre‐core/core; X protein |
L‐6T |
Abbreviations: PCMTD2, protein‐L‐isoaspartate (D‐aspartate) O‐methyltransferase domain containing 2; GPR116, G‐protein‐coupled receptor 116; ME1, malic enzyme 1.
Chronic infection with HBV is associated with increased risk for development of cirrhosis and HCC. However, the relationship between HBV‐DNA integration and cirrhosis remains uncertain and controversial. Of the 23 early‐onset HCC with detected HBV integration, only 56.5% (12 of 23) have cirrhosis, indicating that HBV integration often occurs in patients without cirrhosis.
HBV Preferentially Integrates Into 8p11 and 8q24 in Early‐Onset HCC
Circos plots were employed to visualize and analyze HBV integration patterns at the chromosomal level in early‐ and late‐onset HCC (Fig. 3A,B). Although the detected HBV integration breakpoints in early‐ and late‐onset HCC were both distributed across the entire genome, the hotspots of integration were different (Fig. 3C). Of note, we discovered that 29.2% (21 of 72) of integration sites in early‐onset HCC were localized on chromosome 8. These integration sites are mainly located in 8p11 (5 of 21; 23.8%), a frequent target of genetic alterations in a wide variety of human cancers, including HCC and breast cancer,25 and the 8q24 gene desert (8 of 21; 38.1%), a region involved in susceptibility to various cancers27 (Fig. 4A; Supporting Tables 1 and 2).
Figure 3: Distribution of HBV integration breakpoints on early‐ (A) and later‐onset (B) HCC genome. In this CIRCOS figure, we show the human chromosomes and the integration frequency of each breakpoint on the level of supporting reads as well as on the level of samples. The outer circle represents 24 chromosomes with different color codes and numbers labeled. Each bar represents the frequency of HBV integration breakpoints at a particular locus in the human genome (hg19). Red and blue bars correspond to the HBV integration events at the RefGene (exons, introns, and promoter) and intergenic regions, respectively.
Genes near these breakpoints in the human genome were annotated using the University of California San Francisco (UCSC) Genome Browser (GRCh37/hg19). The genic region was defined as the combination of promoters (5 kilobases [kb] upstream of the transcription start sites, exons, introns, and 3' untranslated region), which is susceptible to the activation of proto‐oncogene expression or inactivation of tumor‐suppressor genes.28 Enrichment of the integration sites in the genic region was observed in 73.8% (62 of 84) in the late‐onset group and 54.2% (39 of 72) in the early‐onset group, suggesting that HBV integration mediated cis‐activation of proto‐oncogene or inactivation of tumor‐suppressor genes might play an important role in the development of HCC (Fig. 4B).
Figure 4: (A) Distribution of HBV integration breakpoints on human chromosomes. (B) Pie charts show the proportion of HBV integration sites in genic and intergenic regions in the early‐ and late‐onset groups. (C) Proportion of integration breakpoints at human repeat regions and no human repeat regions. Human repeat sequence was annotated using the RepeatMasker program in the UCSC Genome Browser (GRCh37/hg19).
We found that 59.7% (43 of 72) and 53.5% (45 of 84) of integration sites in early‐ and late‐onset HCC, respectively, are located in human repeat sequences, such as long interspersed nuclear elements (LINE), short interspersed nuclear elements (SINE; which includes the Alu family), or simple repeats (microsatellites; Fig. 4C). Most of the integration sites located in human repeat sequences do not occur in genic regions; therefore, methods based on Alu PCR will not detect numerous HBV integration sites with important functions.
A handful of recurrent hotspots were observed in both early‐ and late‐onset HCC. Of them, fibronectin 1 (FN1), myeloid/lymphoid or mixed‐lineage leukemia 4 (MLL4), and telomerase reverse transcriptase (TERT) have been previously reported on, whereas synaptotagmin XII (SYT12) and glycine amidinotransferase (GATM), an intergenic site between C‐MYC and PVT1 in early‐onset HCC, and signal transducer and activator of transcription 1 (STAT1), albumin (ALB), suppression of tumorigenicity 18 (ST18), and syntaxin binding protein 4 STXBP4 in late‐onset HCC have not been reported on before (Table 2). The most noteworthy observation is that 20% (6 of 30) of tumors in the early‐onset group had HBV integrated sites in human chromosome 8q24, a large gene desert susceptible to chromosomal translocation and viral integration.27 More interestingly, five of the integration sites located between c‐Myc and PVT1 have the exact viral‐human flanking sequence (Fig. 6B). However, this integration site was not found in tumor samples from the late‐onset group. The integrate site was then examined in all the 113 early‐onset and 145 late‐onset HCC samples. In total, we found 12.4% (14 of 113) in early‐onset HCC, but only 1.4% (2 of 145) in late‐onset HCC have this integration site (P < 0.0001).
Integration Sites in the HBV Genome
We also investigated the characteristics of HBV integration breakpoints in the HBV genome (Fig. 5). Among these integration sites, 32.8% (25 of 76, early‐onset group) and 31.0% (26 of 84, late‐onset group) were significantly enriched in a 400‐bp region of the HBV genome from 1,600 to 2,000 bp, close to direct repeat 1 (DR1) and direct repeat 2 (DR2); (DR2), around the 3'‐end of the HBV x gene (HBx) and 5'‐end of the pre‐C/C genes. Consistent with the observation that higher proportion of integration sites occurring at the 3' of HBx gene, these findings were consistent with the results of the WGS study.29 Interestingly, our results also showed that 40.8% (31 of 76) of the breakpoints were located in the pre‐S region (400‐800 bp) in the early‐onset group, significantly higher than the 26.2% (22 of 84) in the late‐onset group (P < 0.001), which has not been reported on previously.
Figure 5: Distribution of HBV integration breakpoints on the HBV genome. Frequency of integration breakpoints at different loci in the HBV genome (NC_003977) is shown. Locations of the genes encoding HBV polymerase, core, S, and X proteins are shown. Genomic positions are numbered.
Figure 6: Influence of HBV integration on gene expression in HCC. HBV breakpoint sites on the recurrently affected genes were mapped to the human hg19 reference sequence. Each read ball represents the location of an HBV breakpoint identified. Chr, chromosome. Boxes represent exons, and open arrows show orientation of the genes. Expression of ALB (A), C‐Myc, and PVT1 (n = 16; B) in paired tumors and normal samples was analyzed by real‐time RT‐PCR and normalized to glyceraldehyde 3‐phosphate dehydrogenase expression. Relative levels of miR‐1204 expression were normalized to Hs_RNU6B. Data shown are mean ± standard error of the mean of three separate experiments.
Influence of HBV Integration on Gene Expression in HCC
We next investigated the possible cis‐activating effect of HBV‐DNA integration hotspots on expression of adjacent cellular genes. Compared with samples without these integration sites, HBV integration at the TERT promoter, MLL4 intron, and ALB intron increased gene expression in tumor samples, relative to paired nontumor samples (fold change, T/N>2; Fig. 6A and Supporting Fig. 1A,B).
In the early‐onset group, the recurrent hotspot located in 8q24 is embedded in a SINE repeat sequence (chr8:128764676‐128764961), which is approximately 11 kb downstream of c‐Myc and 30 kb upstream of PVT1. Previous studies have reported that viral integration and human chromosomal translocations can disrupt the expression of genes over hundreds of kilobases away. Therefore, we investigated the expression of c‐MYC, PVT1, and miR‐1204 in tumor samples, relative to paired nontumor samples by real‐time reverse‐transcription (RT)‐PCR analysis. Expression of c‐MYC and PVT1 and miR‐1204 in the 16 samples having the integration site were greatly higher than 7 tumors without this integration site (P < 0.001; Fig. 6B). Therefore, HBV integration into this site potentially contributes to the development of early‐onset HCC, which has not been previously reported on.
Discussion
Early‐onset HCC accounts for 15%‐20% of total HCCs in Asia, and incidence is increasing. Although HBV infection is the most epidemiologically associated risk factor for early‐onset HCC, its role in hepatocarcinogenesis remains unclear. In the present study, we conducted a comparative analysis of the HBV subgenotype and integration in early‐onset (≤30) and late‐onset (≥70) HCC patients. To the best of our knowledge, this is the first intensive investigation of the role of HBV infection in early‐onset HCC. The present study may help to clarify HCC risk factors in young HBV carriers.
Some data associate genotype B with development of HCC in young HBV carriers. A 15‐year study of 460 carriers of HBV reported that genotype B was the most frequent genotype among 26 children with HBV‐related HCC (found in 74%),31 indicating that the HBV subgenotype might contribute to early‐onset HCC. Of note, the HBV subgenotype B2 distribution is significantly different in early‐ and late‐onset HCC (47.8% vs. 26.2%, respectively; P < 0.001) in our cohort. An epidemiological study with a large cohort including 81,775 patients from various geographical regions in mainland China4 reported that the predominant HBV genotypes found in Eastern China are B2 (25.7%) and C2 (66.2%). Given that the patients in our cohort are also from the same geographical region, the HBV subgenotype B2 is increased in early‐onset HCC, compared with the general distribution of HBV B2 in our geographical area (47.8% vs. 25.7%), indicating that the B2 subgenotype may serve as a risk factor in early‐onset HCC patients. This finding is consistent with the data from Taiwan and Mainland China, regions where genotype B was found to be predominant in younger HCC patients, at 90% in those <35 years of age and most without cirrhosis.16
A significant feature of chronic HBV infection is the integration of subgenomic HBV‐DNA fragments into different locations within the host DNA.32 Furthermore, chronic infection with HBV is associated with increased risk for development of cirrhosis and HCC. However, the role of HBV integration in HCC development remains uncertain. Theoretically, HBV integration is increased in late‐onset HCC given the increased time of HBV infection. Unexpectedly, a high frequency of HBV integration is reported to be more common in early‐onset HCC than in late‐onset HCC.33 We also observed that HBV integration is a common event in both early‐ and late‐onset HCC. Although the reason for the high‐frequency integration in early‐onset patients with HCC is unclear, these findings indicated that HBV integration might also play an important role in HCC development.
Previous studies have suggested that HBV DNA is randomly integrated into the human genome. However, it is important to note that these initial studies were limited by technological constraints to the analysis of only certain portions of the host genome and were also constrained by small sample sizes. Recent studies using the new technology of next‐generation sequencing (NGS) have suggested that HBV integration might accumulate at specific sites in HCC, such as the gene encoding human TERT, MLL4, cyclin E1, SUMO1/sentrin‐specific peptidase 1, and rho‐associated, coiled‐coil containing protein kinase 1.29 Our results also support the notion that HBV integration is not randomly distributed in the human genome. First, we found several recurrent hotspots both in early‐ (SYT12, GATM, and FN1) and late‐onset HCC (STAT1, ALB, MLL4, and TERT). MLL4, TERT, and FN1 have been previously reported as HBV integration hotspots.16 Second, HBV integration often occurs in the human repeat locus, such as LINE, SINE, or microsatellites; these regions are not randomly distributed in the human genome.
The most noteworthy observation is the recurrent hotspot in early‐onset HCC mainly located in 8p11 and 8q24. The short arm of chromosome 8p11‐12 is a frequent target of genetic alterations in a wide variety of human cancers.25 No oncogene from the 8p11‐12 amplification unit has been identified with certainty.26 A recent study by Lau et al.35 reported a recurrent HBV integration hotspot in HBV‐associated HCC, which is located in a LINE1 repeat region in Chr. 8p11 (chr 41,493,923‐41,494,302). This integration site was detected in 21 of 90 HCCs (23.3%). In all cases, the exact viral‐human flanking sequence was found. HBV integration into this site produces an oncogenic HBV‐LINE1 chimeric transcript that activates Wnt signaling and correlates with poorer patient survival.
The 8q24 region has been described as a “gene desert” because of the paucity of functionally annotated genes located within this region.36 However, genome‐wide association studies have identified a large number of single‐nucleotide polymorphisms in chr.8q24 that are linked to susceptibility for different diseases, including cancers of the prostate, breast, esophagus, head and neck, ovarian, colon, and pancreas.37 Interestingly, we also detected a hotspot in 8q24 (between c‐MYC and PVT1) in early‐onset HCC. We found that this integration site has a similar characteristic with the site reported by Lau et al.: They found that HBV integrated into LINE1 and the 5' HBV sequence shared a GTG sequence with the 3' end of the human LINE1. In our cohort, the integration site between c‐MYC and PVT1 is also located in a human SINE sequence (chr8:128764676‐128764961); in addition, 5' HBV sequence also shared a 3‐bp microhomogy sequence (CCA) with the 3' end of the human SINE. These findings provided evidence that HBV integration is not randomly distributed in the human genome, and the mechanism of HBV insertion into the host genome was likely generated by the mediation of microhomology during break‐induced replication repair.39
We found that HBV integration into this site correlates well with c‐MYC, PVT1, and miR‐1204 expression. PVT1 overexpression promotes tumorigenesis by inhibiting cell apoptosis in ovarian and breast cancer.40 In addition, PVT1 is a noncoding host gene that encodes six microRNAs (miR‐1204, −1205, −1206, 1207‐5p, miR‐1207‐3p, and −1208). miR‐1204 has been reported to be a candidate oncogene that enhances oncogenesis in combination with MYC.41 We found that this integration site was detected in 14 of the 113 early‐onset HCCs, whereas it was detected in only 2 of the 145 late‐onset HCC, strongly suggesting that HBV integration into this site potentially contributes to development of early‐onset HCC, which has not been previously reported on.
Several limitations in our study should be addressed. Because next‐generation sequencing remains expensive and cost prohibitive for sequencing a large number of samples, we used pooled samples for HIVID analysis; thus, various integrated breakpoints that occur at a low frequency are potentially missed. In addition, although we identified various distinct features in early‐ and late‐onset HCC groups, these findings need to be further investigated in a large cohort given the limited sample size. It remains unclear at which time in tumor development this integration event occurs and whether it plays an important, causative role in HCC development or progression; therefore, further studies are needed to elucidate the role of this integration site in HCC development or progression.
References
1. Center MM, Jemal A. International trends in liver cancer incidence rates. Cancer Epidemiol Biomarkers Prev 2011;20:2362‐2368.
2. Ishikawa T. Clinical features of hepatitis B virus‐related hepatocellular carcinoma. World J Gastroenterol 2010;16:2463‐2467.
3. El‐Serag HB, Mason AC. Rising incidence of hepatocellular carcinoma in the United States. N Engl J Med 1999;340:745‐750.
4. Yin J, Zhang H, He Y, Xie J, Liu S, Chang W, et al. Distribution and hepatocellular carcinoma‐related viral properties of hepatitis B virus genotypes in Mainland China: a community‐based study. Cancer Epidemiol Biomarkers Prev 2010;19:777‐786.
5. Bruix J, Sherman M; American Association for the Study of Liver Diseases . Management of hepatocellular carcinoma: an update. Hepatology 2011;53:1020‐1022.
6. Sezaki H, Kobayashi M, Hosaka T, Someya T, Akuta N, Suzuki F, et al. Hepatocellular carcinoma in noncirrhotic young adult patients with chronic hepatitis B viral infection. J Gastroenterol 2004;39:550‐556.
7. Kim JH, Choi MS, Lee H, Kim do Y, Lee JH, Koh KC, et al. Clinical features and prognosis of hepatocellular carcinoma in young patients from a hepatitis B‐endemic area. J Gastroenterol Hepatol 2006;21:588‐594.
8. Yang HI, Lu SN, Liaw YF, You SL, Sun CA, Wang LY, et al. Hepatitis B e antigen and the risk of hepatocellular carcinoma. N Engl J Med 2002;347:168‐174.
9. Lam CM, Chan AO, Ho P, Ng IO, Lo CM, Liu CL, et al. Different presentation of hepatitis B‐related hepatocellular carcinoma in a cohort of 1863 young and old patients ‐ implications for screening. Aliment Pharmacol Ther 2004;19:771‐777.
10. Namieno T, Kawata A, Sato N, Kondo Y, Uchino J. Age‐related, different clinicopathologic features of hepatocellular carcinoma patients. Ann Surg 1995;221:308‐314.
11. Livingston SE, Simonetti JP, Bulkow LR, Homan CE, Snowball MM, Cagle HH, et al. Clearance of hepatitis B e antigen in patients with chronic hepatitis B and genotypes A, B, C, D, and F. Gastroenterology 2007;133:1452‐1457.
12. Kao JH, Chen PJ, Lai MY, Chen DS. Hepatitis B genotypes correlate with clinical outcomes in patients with chronic hepatitis B. Gastroenterology 2000;118:554‐559.
13. Yuan J, Zhou B, Tanaka Y, Kurbanov F, Orito E, Gong Z, et al. Hepatitis B virus (HBV) genotypes/subgenotypes in China: mutations in core promoter and precore/core and their clinical implications. J Clin Virol 2007;39:87‐93.
14. Livingston SE, Simonetti JP, McMahon BJ, Bulkow LR, Hurlburt KJ, Homan CE, et al. Hepatitis B virus genotypes in Alaska Native people with hepatocellular carcinoma: preponderance of genotype F. J Infect Dis 2007;195:5‐11.
15. Farazi PA, DePinho RA. Hepatocellular carcinoma pathogenesis: from genes to environment. Nat Rev Cancer 2006;6:674‐687.
16. Paterlini‐Brechot P, Saigo K, Murakami Y, Chami M, Gozuacik D, Mugnier C, et al. Hepatitis B virus‐related insertional mutagenesis occurs frequently in human liver cancers and recurrently targets human telomerase gene. Oncogene 2003;22:3911‐3916.
17. Saigo K, Yoshida K, Ikeda R, Sakamoto Y, Murakami Y, Urashima T, et al. Integration of hepatitis B virus DNA into the myeloid/lymphoid or mixed‐lineage leukemia (MLL4) gene and rearrangements of MLL4 in human hepatocellular carcinoma. Hum Mutat 2008;29:703‐708.
18. Tamori A, Yamanishi Y, Kawashima S, Kanehisa M, Enomoto M, Tanaka H, et al. Alteration of gene expression in human hepatocellular carcinoma with integrated hepatitis B virus DNA. Clin Cancer Res 2005;11:5821‐5826.
19. Li W, Zeng X, Lee NP, Liu X, Chen S, Guo B, et al. HIVID: an efficient method to detect HBV integration using low coverage sequencing. Genomics 2013;102:338‐344.
20. Chen J, Yin J, Tan X, Zhang H, Zhang H, Chen B, et al. Improved multiplex‐PCR to identify hepatitis B virus genotypes A‐F and subgenotypes B1, B2, C1 and C2. J Clin Virol 2007;38:238‐243.
21. Yan H, Jin H, Xue G, Mei Q, Ding F, Hao L, Sun SH. Germline hMSH2 promoter mutation in a Chinese HNPCC kindred: evidence for dual role of LOH. Clin Genet 2007;72:556‐561.
22. Yin J, Zhang H, Li C, Gao C, He Y, Zhai Y, et al. Role of hepatitis B virus genotype mixture, subgenotypes C2 and B2 on hepatocellular carcinoma: compared with chronic hepatitis B and asymptomatic carrier state in the same area. Carcinogenesis 2008;29:1685‐1691.
23. Fujie H, Moriya K, Shintani Y, Yotsuyanagi H, Iino S, Koike K. Hepatitis B virus genotypes and hepatocellular carcinoma in Japan. Gastroenterology 2001;120:1564‐1565.
24. Chan HL, Hui AY, Wong ML, Tse AM, Hung LC, Wong VW, Sung JJ. Genotype C hepatitis B virus infection is associated with an increased risk of hepatocellular carcinoma. Gut 2004;53:1494‐1498.
25. Emi M, Fujiwara Y, Nakajima T, Tsuchiya E, Tsuda H, Hirohashi S, et al. Frequent loss of heterozygosity for loci on chromosome 8p in hepatocellular carcinoma, colorectal cancer, and lung cancer. Cancer Res 1992;52:5368‐5372.
26. Gelsi‐Boyer V, Orsetti B, Cervera N, Finetti P, Sircoulomb F, Rouge C, et al. Comprehensive profiling of 8p11‐12 amplification in breast cancer. Mol Cancer Res 2005;3:655‐667.
27. Huppi K, Pitt JJ, Wahlberg BM, Caplen NJ. The 8q24 gene desert: an oasis of non‐coding transcriptional activity. Front Genet 2012;3:69.
28. Toh ST, Jin Y, Liu L, Wang J, Babrzadeh F, Gharizadeh B, et al. Deep sequencing of the hepatitis B virus in hepatocellular carcinoma patients reveals enriched integration events, structural alterations and sequence variations. Carcinogenesis 2013;34:787‐798.
29. Sung WK, Zheng H, Li S, Chen R, Liu X, Li Y, et al. Genome‐wide survey of recurrent HBV integration in hepatocellular carcinoma. Nat Genet 2012;44:765‐769.
30. Jiang Z, Jhunjhunwala S, Liu J, Haverty PM, Kennemer MI, Guan Y, et al. The effects of hepatitis B virus integration into the genomes of hepatocellular carcinoma patients. Genome Res 2012;22:593‐601.
31. Ni YH, Chang MH, Wang KJ, Hsu HY, Chen HL, Kao JH, et al. Clinical relevance of hepatitis B virus genotype in children with chronic infection and hepatocellular carcinoma. Gastroenterology 2004;127:1733‐1738.
32. Bonilla Guerrero R, Roberts LR. The role of hepatitis B virus integrations in the pathogenesis of human hepatocellular carcinoma. J Hepatol 2005;42:760‐777.
33. Chang MH, Chen PJ, Chen JY, Lai MY, Hsu HC, Lian DC, et al. Hepatitis B virus integration in hepatitis B virus‐related hepatocellular carcinoma in childhood. Hepatology 1991;13:316‐320.
34. Murakami Y, Saigo K, Takashima H, Minami M, Okanoue T, Brechot C, Paterlini‐Brechot P. Large scaled analysis of hepatitis B virus (HBV) DNA integration in HBV related hepatocellular carcinomas. Gut 2005;54:1162‐1168.
35. Lau CC, Sun T, Ching AK, He M, Li JW, Wong AM, et al. Viral‐human chimeric transcript predisposes risk to liver cancer development and progression. Cancer Cell 2014;25:335‐349.
36. Wasserman NF, Aneas I, Nobrega MA. An 8q24 gene desert variant associated with prostate cancer risk confers differential in vivo activity to a MYC enhancer. Genome Res 2010;20:1191‐1197.
37. Easton DF, Eeles RA. Genome‐wide association studies in cancer. Hum Mol Genet 2008;17:R109‐R115.
38. Grisanzio C, Freedman ML. Chromosome 8q24‐associated cancers and MYC. Genes Cancer 2010;1:555‐559.
39. Hastings PJ, Ira G, Lupski JR. A microhomology‐mediated break‐induced replication model for the origin of human copy number variation. PLoS Genet 2009;5:e1000327.
40. Guan Y, Kuo WL, Stilwell JL, Takano H, Lapuk AV, Fridlyand J, et al. Amplification of PVT1 contributes to the pathophysiology of ovarian and breast cancer. Clin Cancer Res 2007;13:5745‐5755.
41. Shtivelman E, Bishop JM. The PVT gene frequently amplifies with MYC in tumor cells. Mol Cell Biol 1989;9:1148‐1154.
42. Carramusa L, Contino F, Ferro A, Minafra L, Perconti G, Giallongo A, Feo S. The PVT‐1 oncogene is a Myc protein target that is overexpressed in transformed cells. J Cell Physiol 2007;213:511‐518.
