The authors have declared that no competing interests exist.
Conceived and designed the experiments: ESH YHK BJK. Performed the experiments: DWK SAL. Analyzed the data: DWK BJK. Contributed reagents/materials/analysis tools: DWK SAL. Wrote the paper: DWK BJK. Obtained permission for use of plasma serum: DWK.
Previous studies have proved the presence of several distinct types of mutations in hepatitis B virus (HBV) infections, which are related to the progression of liver disease. However, few reports have detailed the mutation frequencies and mutation patterns in the precore/core (preC/C) region, which are based on the clinical status and HBeAg serostatus. Our aim in this study is to investigate the relationships between the preC/C mutations and clinical severity or HBeAg serostatus from patients chronically infected with HBV genotype C. A total of 70 Korean chronic patients, including 35 with hepatocellular carcinoma (HCC), participated in this study. HBV genotyping and precore/core mutations were analyzed by direct sequencing. All patients were confirmed to have genotype C infections. Mutations in the C region were distributed in a non-random manner. In particular, mutations in the MHC class II restricted region were found to be significantly related to HCC. Six (preC-W28*, C-P5H/L/T, C-E83D, C-I97F/L, C-L100I and C-Q182K/*) and seven types (preC-W28*, preC-G29D, C-D32N/H, C-E43K, C-P50A/H/Y, C-A131G/N/P and C-S181H/P) of mutations in the preC/C region were found to be related to HCC and to affect the HBeAg serostatus, respectively. In conclusion, our data indicated that HBV variants in the C region, particularly in the MHC class II restricted region, may contribute to the progress of HCC in chronic patients infected with genotype C. In addition, we found several distinct preC/C mutations in the Korean chronic cohort, which affect the clinical status of HCC and HBeAg serostatus of patients infected with genotype C.
Despite the availability of an effective vaccine, more than 350 million people worldwide are chronically infected with the hepatitis B virus (HBV), and many people have developed serious liver diseases, such as cirrhosis and hepatocellular carcinoma (HCC)
Based on intergroup divergence of >8% in the complete genome sequence, hepatitis B virus (HBV) strains are classified into eight genomic groups or genotypes, which are designated as A–H
Over the past decade, increasing attention has been focused on variant HBV strains that contribute to the clinical severity of liver diseases, especially HCC. To date, a number of mutation patterns of HBV, such as the precore mutation at nucleotide 1896 (G→A) or the double mutation in the basal core promoter (BCP) region at nucleotides 1762 (A→T) and 1764 (G→A), have been widely studied as HBV mutations related to clinical severity
The HBV C protein (HBcAg), the protein shell of the virus core, is 183 residues long, of which 149 residues of the N-terminal are the assembly domain
Relationships between the frequencies of the preC/C region and the progression of liver disease have been elucidated
Plasma serum samples were collected from 70 chronic hepatitis B patients who visited the Seoul National University Hospital in 2005. Among these, 35 serum samples were HBeAg-positive and 35 were HBeAg-negative. Clinical diagnoses of the subjects were chronic hepatitis (n = 27), liver cirrhosis (n = 8) and HCC (n = 35). Chronic liver disease definitions were as follows: chronic hepatitis was defined as an elevation or fluctuation of serum ALT over 6 months without any evidence of any other chronic liver disease
200
To analyze the mutation patterns and their frequencies of deletions and insertions in the entire preC/C region, a nested PCR protocol was used. First-round PCR was performed using the sense primer CoreF1 (
The purified PCR products were directly determined using both primers of the secondary PCR, CoreF2 and CoreR2, and with a dideoxy method that uses a BigDye Terminator Cycle Sequencing Ready reaction, V. 2, and a fluorescent 373A DNA sequencer (Applied Biosystems, Foster City, CA, USA). If two peaks were present at any position within the chromatograph, the dominant peak was utilized.
For genotyping, a phylogenetic analysis based on entire sequences of the entire preC/C region (639 bp) was performed on all 70 HBV strains. The nucleotide sequences of the 70HBV strains were compared with those of 17 reference strains, representing each of the genotypes of A–D including 14 genotype C strains obtained from GenBank [Accession numbers AB031262 (C), AB074755 (C), AB100695 (B), AF068756 (C), AF223957 (C), AF286594 (C), AY123041 (C), AY247030 (C), AY247032 (C), D12980 (C), D16667 (C), D23680 (C), D23682 (C), GQ377616 (C), X01587 (C), X70185 (A) and X72702 (D)]. A mutation was defined through comparisons with the consensus sequence of HBV strains in our cohort and 17 reference strains. Phylogenetic trees were inferred by means of neighbor-joining with MEGA version 4.0.2
Results were expressed as percentages, means ± SD, or as medians (range). The differences between categorical variables were analyzed using Fisher's exact test or a Chi-square test. For continuous variables, the Student's
The phylogenetic analysis based on all of the 639-bp sequences of the preC/C region indicated that all 70 HBV strains from Korean patients belonged to genotype C, irrespective of their clinical status or HBeAg serostatus (Fig. S1).
Mutations were observed in a total of 82 out of 212 codons (29 codons in preC and 183 codons in C) from 60 patients out of 70 (85.7%) (
Black and grey bars represent mutations related to HCC and mild forms of liver disease, respectively. The dark-shaded regions are the MHC class I-restricted (core aa 18–27, 88–96, 130–140, 141–151) and the light-shaded regions are MHC class II-restricted T-cell epitopes of HBcAg (core aa 1–20, 50–69, 81–105, 117–131, 141–165). Asterisks and triangles indicate specific mutations related to HCC and affecting the HBeAg serostatus, respectively, in the present study. The red arrow indicates the six mutations, which were previously reported to be negatively related to HCC
Clinical factors | Wild type (n = 10) | Mutation (n = 60) |
|
Age in years, mean ± SD | 36.9±16.1 | 51.9±12.7 | 0.001 |
Male (%) | 7 (70) | 46 (76.7) | N.S |
HBeAg-positive (%) | 6 (60) | 29 (48.3) | N.S |
Liver disease (no.) CH:LC:HCCC | 8∶1∶1 | 19∶7∶34 | - |
ALT status (%) |
6 (60) | 47 (81) | N.S |
HBV-DNA (pg/ml) median (range) | 6.38E+06 (0–5.50E+07) | 1.89E+06 (0–8.24E+07) | N.S |
A mutation was defined through comparisons with the consensus sequence of HBV strains in our cohort and 17 reference strains.
N.S: Not significant.
CH: Chronic hepatitis, LC: Liver cirrhosis, HCC: Hepatocellular carcinoma.
The number of patients whose ALT levels were greater than the upper limits of normal ALT levels for men (30 IU/L) and women (19 IU/L)
Generally, a non-random distribution between the mutations of immuno-active region and immuno-inactive regions was shown. The mutation rates in the immuno-active regions (MHC class I + II) were significantly higher than those of the immuno-inactive region (2.2% vs. 1.7%, p = 0.016). The mutation rates in the MHC class II restricted region (designated as M2RR), but not in the MHC class I restricted region (designated M1RR), were significantly higher than those in the immuno-inactive region (2.3% vs. 1.7%, p = 0.009). This contrast was more pronounced considering the mutation rates (4.1%) in “hot spots”, aa residue 81–105 region in M2RR (4.1% vs. 1.7%, p<0.001) (
Region (n = 14,840) | No. of mutations/No. of codons | Mutation rate (%) | |
Immuno-inactive regions | 107/ 6,370 | 1.7 | - |
Immuno-active regions | 190/8,470 | 2.2 | 0.016 |
- MHC class I epitope | 61/2,870 | 2.1 | N.S |
- MHC class II epitope | 170/7,350 | 2.3 | 0.009 |
- 81-105 region in MHC class II | 72/1,750 | 4.1 | <0.001 |
P-values were determined by a comparison with the mutation rate of immuno-inactive regions.
N.S: Not significant.
Statistically significant after a Bonferroni post hoc analysis (p<0.05).
Immuno-active regions include MHC class I-restricted epitopes, MHC class II-restricted epitopes, and the 81–105 region of MHC class II.
As expected, mutation in the 28th codon (tryptophan to stop, designated preC-W28*), previously known as the “mutational hot spot” (1896 preC mutation) leading to the inhibition of HBeAg production and related to the progression of liver diseases
The mutation rates of the entire preC/C region in HCC patients (2.2%) tended to be higher than in the comparison group, patients with liver cirrhosis (LC) and those with chronic hepatitis (CH) (1.8%) (p = 0.061). Mutation rates in immuno-active regions, but not in immune-inactive regions were significantly higher in HCC patients than in the comparison group [HCC (2.6%) vs. the comparison group (1.9%), p = 0.033]. Of these, the mutation rates in M2RR (2.7% vs. 1.9%, p = 0.024), but not in M1RR (2.4% vs. 1.8%, p = 0.3), were significantly higher in HCC patients than in comparison group. Furthermore, their difference in the aa residue 81–105 region was also more pronounced (5.6% vs. 2.6%, p = 0.002) (
Classification | HCC (%)/No. of codons | LC+CH (%)/No. of codons | |
Immuno-inactive regions | 55 (1.7)/3,185 | 52 (1.6)/3,185 | N.S |
Immuno-active regions | 110 (2.6)/4,235 | 80 (1.9)/4,235 | 0.033 |
- MHC class I epitope | 35 (2.4)/1,435 | 26 (1.8)/1,435 | N.S |
- MHC class II epitope | 100 (2.7)/3,675 | 70 (1.9)/3,675 | 0.024 |
- 81–105 region in MHC class II | 49 (5.6)/875 | 23 (2.6)/875 | 0.002 |
Total | 165 (2.2)/7,420 | 132 (1.8)/7,420 | 0.061 |
N.S: Not significant.
Statistically significant after a Bonferroni post hoc analysis (p<0.05).
Overall, the mutation frequencies in the entire preC/C region in the HBeAg-negative groups (2.5%) were significantly higher than in the HBeAg-positive groups (1.5%) (p<0.001). However, some differences according to respective regions within HBcAg were found. The differences in the mutation rates in M1RR between the two groups (2.6% vs. 1.7%, p = 0.094) did not reach a statistically significant level, However, the mutation rates of M2RR (3.0% vs. 1.7%, p<0.001) were significantly higher in HBeAg-negative patients as compared to HBeAg-positive patients (
Mutation rate | HBe− (%)/No. of codons | HBe+ (%)/No. of codons | |
Immuno-inactive regions | 67 (2.1)/3,185 | 40 (1.3)/3,185 | 0.011 |
Immuno-active regions | 117 (2.8)/4,235 | 73 (1.7)/4,235 | 0.002 |
- MHC class I epitope | 37 (2.6)/1,435 | 24 (1.7)/1,435 | 0.094 |
- MHC class II epitope | 109 (3.0)/3,675 | 61 (1.7)/3,675 | <0.001 |
- 81–105 region in MHC class II | 44 (5.0)/875 | 28 (3.2)/875 | 0.07 |
Total | 184 (2.5)/7,420 | 113 (1.5)/7,420 | <0.001 |
Statistically significant after a Bonferroni post hoc analysis (p<0.05).
Five mutations in the C region (C-P5H/L/T, C-E83D, C-I97F/L, C-L100I and C-Q182K/*) and one in preC (preC-W28*) were found to be related to HCC patients compared to patients at other stages of the disease, such as LC and CH, respectively. Generally, five mutations in the C region were found at a significantly higher level in HCC patients than in patients at other stages of the disease. However, the predominance of preC-W28* in HCC patients did not reach a statistically relevant level (p = 0.093) (
The types of mutations within the region of MHC I and II-restricted T-cell epitopes are marked.
Five mutations in the C region (C-D32N/H, C-E43K, C-P50A/H/Y, C-A131G/N/P and C-S181H/P) and two in preC (preC-W28* and preC-G29D) were found to affect the HBeAg serostatus (
Our data showed that the majority of patients (60/70 patients, 85.7%) had more than one mutation in the preC/C region (
Our data link four major factors, old age, the location of the T cell epitope (particularly M2RR), the HBeAg negative serostatus, and the HCC clinical status, to higher mutation rates in the preC/C region. First, the positive relationship between the preC/C mutation frequency and old age [wild type (36.9) vs. mutation (51.9), p = 0.001] suggests that the accumulation of preC/C mutations during the natural course of CHB contributes to the persistent infection of HBV in areas where vertical infection is predominant (
Second, preC/C mutations in our cohort were distributed in a non-random manner, as shown in other studies (
Third, although differences in clinical factors apart from ALT level between patients with two different HBeAg serostatus were not found (Table S3), a significant correlation between the HBeAg-negative serostatus and preC/C mutations was observed in this study [HBe- (2.5%) vs. HBe+ (1.5%), p<0.001] (
Finally, our data showed significant relationships between M2RR preC/C mutations, though not the immuno-inactive region or M1RR mutations, and HCC patients [HCC (2.7%) vs. comparison group (1.9%), p = 0.024] (
Of seven mutations affecting HBeAg serostatus observed in this study, it is noteworthy that five (C-D32N/H, C-E43K, C-P50A/H/Y, C-A131G/N/P and C-S181H/P) in the C region may have been first introduced in relation to the HBeAg serostatus. In particular, to the best of our knowledge, two types of mutations (C-D32N/H, and C-E43K) related to the HBeAg positive serostatus have not been introduced to date in HBV variants. Interestingly, both mutations were not located at the regions of T or B cell epitopes, suggesting that their mutations are induced by other mechanisms, rather than by immune evasion. Generally, mutations associated with the HBeAg-negative serostatus are known to have the potential to be related to disease severity
Currently, relationships between the frequencies of the C region and the progression of liver disease have been assessed
C-I97F/L, most frequently found in HCC-related mutations, is well known as the most frequently encountered HBcAg mutation, as mentioned in several studies
All of our sequence data were based on the direct sequencing protocol, which has the potential to underestimate deletions or point mutations at a level of less than 50% in each patient's viral quasispecies. Therefore, to resolve this concern, molecular based approaches such as real-time PCR (RT-PCR) based on fluorescence-resonance energy transfer (FRET) technology
Of note, high frequency of the lower level of HBV DNA in mutated strains than in wild strains were found in our cohort (
In conclusion, our data indicated that HBV variants in the C region, particularly in M2RR, may contribute to HCC progress in chronic patients infected with the genotype C via immune evasion against the CD4 T cell-mediated immune response. Furthermore, the identified HBcAg mutations related to HCC and affecting HBeAg serostatus have the potential to serve as diagnostic markers to detect early on the progression of liver disease, including HCC.
A phylogenic tree based on the sequence of the preC/C region from 70 and nine reference HBV strains. Genetic distances were estimated using the Kimura two-parameter matrix and the phylogenetic tree was constructed using the neighbor-joining method. The percentages indicated at the nodes represent bootstrap levels supported by 1000 re-sampled data sets. Bootstrap values of less than 50% are not shown. The bar indicates 1% estimated sequence divergence.
(TIF)
Comparison of clinical features according to liver diseases.
(DOC)
Comparison of the Clinical Features between patients with wild type and I97F/L.
(DOC)
Comparison of Clinical Features of Patients according to HBeAg Serostatus.
(DOC)