Potential conflict of interest: Nothing to report.
Mannose binding lectin (MBL) is a pattern-recognition molecule of the innate immune system. The roles of MBL and its gene (mbl2) polymorphisms, −221X/Y and codon 54A/B, in hepatitis B virus (HBV) infection were investigated in this study. We recruited 320 nonprogressed hepatitis B surface antigen (HBsAg) carriers; 199 progressed HBsAg carriers with hepatocellular carcinoma or cirrhosis; 87 spontaneously recovered individuals who were HBsAg negative and anti-HBs and anti HBc positive; and 484 controls who were naïve to HBV. There was no significant difference between nonprogressed carriers, spontaneously recovered individuals, and controls in terms of serum MBL levels and mbl2 polymorphisms distributions. However, the low MBL genotypes had a dose-dependent correlation with the cirrhosis and hepatocellular carcinoma in progressed carriers with odds ratios of 1.36 and 3.21 for the low and extremely low MBL genotypes, respectively (P = .01). The low-expression promoter haplotype XA (OR = 1.97) and the mutant haplotype YB (OR = 1.90) were also associated with the cirrhosis and hepatocellular carcinoma (P = .002). As expected, the lower serum MBL levels in progressed carriers as compared with nonprogressed carriers were due to an overrepresentation of low and extremely low MBL genotypes. Moreover, MBL could bind HBsAg in a dose- and calcium-dependent and mannan-inhibitable manner in vitro, suggesting that binding occurs via the carbohydrate recognition domains. This binding also enhanced C4 deposition. In conclusion, these results suggest that low MBL genotypes associate with the occurrence of cirrhosis and hepatocellular carcinoma in progressed HBsAg carriers, and MBL can bind HBsAg. (HEPATOLOGY 2005.)
Hepatitis B virus (HBV) infection is one of the major infectious diseases, with more than 350 million carriers worldwide.1, 2 It is the most common cause of acute hepatitis and may progress to chronic liver disease, including cirrhosis or hepatocellular carcinoma.1, 2 The mechanism for this progression is still undetermined. The clinical outcome of HBV infection is highly variable, and genetic factors are likely to affect disease progression after HBV infection.3 Strong evidence supporting a role for genetic factors comes from a study of twins in Taiwan,4 which demonstrated a higher concordance rate for hepatitis B surface antigen (HBsAg) status in monozygotic than in dizygotic twins.
HBV infection and its clinical outcome involve strong genetic components and a number of candidate genes have been discovered. The association of HLA polymorphisms with the outcome of HBV infection has been consistently observed in various populations.5, 6 For example, the HLA-DR13 is associated with the spontaneous elimination of infection in Koreans, Gambians, and Europeans.6 Other genes, such as those encoding interleukin (IL)-10,7 tumor necrosis factor alpha (TNF-α),8 and the vitamin D receptor,9 also have been reported to be associated with HBV infection.
Mannose binding lectin (MBL) is a calcium-dependent C-type lectin with a structural analogy to complement component C1q. It is a key molecule in our innate immunity and functions as an “ante-antibody” in first-line host defence.10 MBL binds through its multiple carbohydrate recognition domains to repeating arrays of carbohydrate structures on microbial surfaces11 and is then able to activate the complement system through specific proteases called MBL-associated serine proteases (MASPs),12 or enhancing phagocytosis by acting as an opsonin.13, 14 Deficiency of MBL may confer a generalized susceptibility to infection in children and adults.15, 16 A risk of infection by specific pathogens in MBL-deficient individuals has also been established, and examples include human immunodeficiency virus, Cryptosporidium parvum, and Neisseria meningitidis.17–19 Serum MBL levels play an important role in regulating the production of inflammatory cytokines such as IL-6, IL-1β, and TNF-α by monocytes in response to pathogen infection,20 and hence may have an impact on disease severity or progression.
MBL gene (mbl2) has two single nucleotide polymorphisms (SNPs) in the promoter (−221 X/Y) and the first exon (codon 54 A/B), which have functional effects on the serum MBL level.21–23 Allele Y is associated with a high-expression promoter when compared with allele X. Allele B of codon 54 refers to a structural mutant allele whereas allele A is the wild type. Because of linkage disequilibrium, the two SNPs only form 3 hapotypes: YA, XA, and YB. YA and XA represent high- and low-expression promoters with wild-type structural MBL, respectively. YB is commonly known to be a mutant haplotype with minimal detectable serum MBL level.24–26 The other two known mutant haplotypes carrying structural polymorphisms, C (codon 57) and D (codon 52), are at an extremely low frequency and probably absent altogether in the Chinese population.21, 22, 27
Several lines of evidence implicate MBL in HBV infection. The preS2 region of the middle hepatitis B surface protein contains a mannose-rich oligosaccharide to which MBL could bind and result in some biological function.28 Polymorphisms at codon 52 and 54 of mbl2 have been demonstrated to associate with chronic HBV infection in some relatively small scale studies,27, 29 although conflicting negative results were also reported.30, 31 Therefore, the role of MBL in HBV infection and the question of whether the mbl2 SNPs are associated with different outcomes of HBV infection deserve further investigation.
We have investigated the binding of MBL to HBsAg and whether this binding would result in complement activation. Also, we studied the association of serum MBL levels and 2 mbl2 SNPs, −221 X/Y and codon 54 A/B, with various outcomes of HBV infection in more than 1,000 Hong Kong Chinese, who were divided into four different groups: progressed HBsAg carriers with cirrhosis or hepatocellular carcinoma, nonprogressed HBsAg carriers, healthy individuals who have spontaneously recovered from HBV infection, and healthy individuals who were naïve to HBV.
HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen; IL, interleukin; TNF, tumor necrosis factor; MBL, mannose binding lectin; MASP, MBL-associated serine protease; SNP, single nucleotide polymorphism; ULN, upper limit of normal; ALT, alanine aminotransferase; IgG, immunoglobulin G; BSA, bovine serum albumin; PBS, phosphate-buffered saline; HBSS, Hank's balanced salt solution; HRP, horseradish peroxidase; HBeAg, hepatitis B e antigen.
Patients and Methods
HBV-Infected Groups and Control Population.
Hong Kong Chinese hepatitis B virus surface antigen (HBsAg) carriers were recruited from Queen Mary Hospital, The University of Hong Kong. The study was approved by the Ethics Committee, Faculty of Medicine, The University of Hong Kong. All carriers provided informed consent, and they were placed into two different groups: the nonprogressed carriers and progressed carriers. Nonprogressed carriers (n= 320, 73.1% male, mean age = 36.5 ± 11.1 years) were HBsAg positive but without any clinical symptoms and were investigated with regular laboratory tests. Of these 320 carriers, 36 (11.3%) had alanine aminotransferase (ALT) > 1 × the upper limit of normal (ULN), and 9 (2.8%) had ALT > 2× ULN. Two hundred twenty-one nonprogressed carriers had platelet counts, of which 207 (93.7%) were normal (>150 × 109/L), and only 6 (2.7%) carriers had platelet counts lower than 100 × 109/L. One hundred twenty-three carriers had ultrasound, of which 115 (93.5%) carriers had no evidence of cirrhosis. Progressed carriers (n = 199, 82.9% male, mean age = 57.9 ± 11.7 years) were HBsAg positive, with chronic liver diseases such as cirrhosis and hepatocellular carcinoma. Progressed carriers with cirrhosis were defined as HBsAg carriers with complications, such as encephalopathy, varices, ascites, or spontaneous bacterial peritonitis, and their regular liver ultrasound scan and liver function test both confirmed cirrhosis. Progressed carriers with hepatocellular carcinoma were defined as HBsAg carriers with a serum alpha-fetoprotein level more than 400 ng/mL combined with typical findings on computed tomography. If these criteria were not met, a histological diagnosis of hepatocellular carcinoma was made by fine-needle aspiration. All HBsAg carriers were further characterized by HBeAg status, HBV DNA titer, and HBV genotype as shown in Table 1.
Table 1. Characterization of Nonprogressed and Progressed Carriers
Nonprogressed Carriers (n = 320)
Progressed Carriers (n = 199)
NOTE. All of the parameters—sex, age, ALT to ULN ratio, HBeAg status, HBV-DNA titer, and HBV genotype—were included in the multivariate analysis as covariables.
P values were calculated with
unpaired t test, and
Mann-Whitney U test.
Six non-progressed and 7 progressed carriers were double infected with HBV genotypes B and C.
Eighty-seven ethnically matched, healthy individuals (66% male, mean age = 29.5 ± 11.7 years) were recruited from Red Cross blood donor centers with antibodies against hepatitis B virus surface antigen (anti-HBs) and anti-hepatitis B virus core antigen (anti-HBc) but HBsAg negative. They were defined as spontaneously recovered individuals. The normal control population consisted of 484 healthy Hong Kong Chinese from Red Cross blood donor centers (56% male, mean age = 24.4 ± 9.5 years) who were HBsAg, anti-HBs, and anti-HBc negative.
Genomic DNA was extracted from EDTA whole blood using a Qiagen DNA Blood Mini kit (Qiagen, CA) according to the manufacturer's instructions. The DNA samples were then stored at 4°C until used.
Genotyping and Haplotyping of mbl2.
Two SNPs in the promoter (−21 X/Y) and the first exon (codon 54 A/B) of mbl2 were genotyped using the TaqMan system (Applied Biosystems, Foster City, CA) as previously described.32
Because of the linkage among SNPs, three haplotypes were deduced: YA, YB, and XA. YA and XA are associated with high- and low-expression promoters, respectively, whereas YB is commonly referred to as a mutant haplotype YB.24–26 We categorized the mbl2 genotypes into 3 subgroups, groups 1, 2, and 3, corresponding to high, low, and extremely low MBL levels, respectively, as described in previous studies.25, 26 Group 1 included YA/YA and YA/XA individuals. Group 2 included XA/XA and YA/YB individuals. Group 3 included the individuals with 2 mutant structural alleles YB/YB or one mutant structural allele with a low expression promoter XA/YB.
Preparation of MBL.
MBL was prepared as previously described.33 Ammonium sulfate was added to 500 g frozen ethanol-fractionated human plasma paste (fraction B+1, equivalent to Cohn fraction I+III; donated by C. Dash, Blood Products Laboratory, Elstree, UK) to give 42% saturation. After dialysis, the solution was applied to a mannan-agarose (Sigma-Aldrich, St. Louis, MO) column (5 mL packed volume; Pharmacia Biotech, Uppsala, Sweden), and 0.01 mol/L EDTA was used to elute the calcium-dependent proteins. The first EDTA eluate was recalcified to 0.05 mol/L CaCl2, and was applied to a second mannan-agarose column and eluted with 0.1 mol/L mannose. The MBL concentration was determined by ELISA. Sample homogeneity was verified by nonreducing SDS-PAGE using a 3% to 10% polyacrylamide gradient gel and silver staining. Bands observed on silver staining were confirmed to be higher-order oligomers of MBL. MBL prepared in this manner is non-covalently associated with MASP.34 In addition, plasma-derived MBL prepared by the State Serum Institute, Copenhagen, Denmark, kindly made available by Dr Claus Koch, was also used in some experiments and gave identical results.
Quantification of Serum MBL Concentrations.
As described previously,32 serum MBL concentration was determined by ELISA. A commercial mouse monoclonal immunoglobulin G (IgG) against human MBL (HYB 131-01, Antibody Shop, Copenhagen, Denmark) was used as the primary antibody, and the same antibody with a biotin label was used as the secondary antibody. Horseradish peroxidase (HRP)-conjugated streptavidin (R&D, Minneapolis, MN) and substrate solution containing tetramethylbenzidine (Substrate Reagent Pack, R&D) were used for signal detection in the assay according to the manufacturer's instructions.
Binding of MBL to HBsAg.
Ninety-six–well flat-bottom polystryrene plates ready-coated with HBsAg (Diagnostic Automation, Calabasas, CA) were used as provided. Alternatively, plates were coated overnight with 1% (w/v) bovine serum albumin (BSA) diluted in phosphate-buffered saline (PBS). The wells were blocked with 1% BSA in PBS with 0.05% NaN3 for 1 hour at room temperature, followed by washing with Hank's balanced salt solution (HBSS) (Gibco-BRL, Grand Island, NY) in 0.05% Tween 20 (HBSS/tw). After washing, the wells were incubated for 2 hours with 100 μL 0, 0.16, 0.31, 0.625, 1.25, 2.5, 5, or 10 μg/mL MBL diluted in HBSS containing 20 mmol/L CaCl2 (HBSS/Ca2+). In some experiments designed to study the characteristics of the MBL-HBsAg binding, wells were incubated with 100 μL MBL diluted in HBSS containing 20 mmol/L EDTA (HBSS/EDTA) or in HBSS/Ca2+ containing 2 mg/mL mannan from Saccharomyces cerevisiae (Sigma-Aldrich, St. Louis, MO) (HBSS/mannan). The wells were washed with HBSS/tw and incubated with 0.2 μg/mL biotinylated monoclonal anti-MBL antibody (HYB 131-01, Antibody Shop) diluted in HBSS/Ca2+ with 0.2% BSA for 1.5 hours at room temperature. HRP-conjugated streptavidin and tetramethylbenzidine substrate solution (R&D) was used for bound antibody detection according to the manufacturer's instructions. The binding of MBL to immobilized HBsAg was evaluated by measuring absolute absorbance values at 450 nm (A450). The HBsAg used for coating was a recombinant HBsAg from a hamster ovary cell line with 97% purity. Its glycosylation pattern was quoted to be 99% to 100% as the native protein (Diagnostic Automation).
Ninety-six–well flat-bottom polystyrene plates ready-coated with HBsAg (Diagnostic Automation) were used as provided. Alternatively, plates were coated overnight with 1% (w/v) BSA diluted in PBS. The wells were blocked with 1% BSA in PBS with 0.05% NaN3 for 1 hour at room temperature. The plate was then washed 3 times with PBS containing 0.05% Tween 20 (PBS/tw). After washing, the wells were incubated with 100 μL of 0.625, 1.25, 2.5, 5, or 10 μg/mL MBL diluted in 20 mmol/L Tris-HCl, 10 mmol/L CaCl2, 1 mol/L NaCl, 0.05% (v/v) Triton X-100, 0.1% BSA, pH 7.4 overnight at 4°C. The wells with no MBL but buffer only served as negative controls. The plates were then washed as described and incubated with 100 μL 4 μg /mL purified human complement component C4 (Quidel, San Diego, CA) diluted in 4 mmol/L barbital, 145 mmol/L NaCl, 2 mmol/L CaCl2, 1 mmol/L MgCl2, 1.5 mmol/L NaN3, pH 7.5, for 1.5 hours at 37°C. The wells were washed 3 times with PBS/tw. Mouse monoclonal anti-C4 antibody (4 μg/mL; Quidel) diluted in PBS/tw was added to the wells for 1.5 hours of incubation at room temperature. The wells were washed as previously described and incubated with 100 μL HRP-anti-mouse IgG (Dako, Glostrup, Denmark), 1/1,000 diluted in PBS with 0.2% BSA, for 1 hour at room temperature. The wells were washed and developed with tetramethylbenzidine substrate solution (R&D) as previously described. The deposition of C4 on HBsAg by MBL pathway was evaluated at A450 according to the following formula: relative percentage of C4 deposition = (A450 − A450 of negative controls) / (A450 of negative controls) × 100.
Statistical analysis was performed using SAS Software, Version 8.02 (Cary, NC). The frequencies of mbl2 genotypes were compared between different groups by overall chi square tests. When significant differences were obtained, logistic regression was used to calculate odds ratios (OR) with 95% confidence intervals (95% CI) and corresponding P values of different genotype frequencies between groups by controlling age, sex, ALT ratio to ULN, HBeAg status, HBV-DNA titer, and HBV genotype as covariables. The most dominant mbl2 genotype or haplotype group was chosen as the reference group (OR = 1). The mbl2 genotypes of all SNPs in the various groups were tested for Hardy-Weinberg equilibrium by chi square tests. Mann-Whitney U tests were used to compare the serum MBL levels and HBV DNA titer between groups. Unpaired t test was used to examine the age difference between the nonprogressed and progressed carriers. In all statistical analyses, a P value of less than .05 was used to reject the null hypothesis.
Serum MBL Levels and the Frequency of mbl2 Polymorphisms Did Not Differ Significantly in Naïve Controls, Spontaneously Recovered Individuals, and Nonprogressed Carriers.
The genotype distributions of mbl2 polymorphisms at −221X/Y and codon 54A/B were compared between the naïve controls, spontaneously recovered individuals and nonprogressed carriers (Table 2). No significant difference was observed. All genotype distributions were consistent with the existence of Hardy-Weinberg equilibrium.
Table 2. mbl2 Genotype and Haplotype in Controls, Spontaneously Recovered Individuals, and Nonprogressed Carriers
Controls (n = 484)
Spontaneously Recovered Individuals (n = 87)
Nonprogressed Carriers (n = 320)
NOTE. The distributions of mbl2 genotype and haplotype were not significantly different among controls, spontaneously recovered individuals, and nonprogressed carriers by overall chi square analysis (P > .05).
YA/YA and YA/XA
YA/YB and XA/XA
YB/XA and YB/YB
The medians (and means) of serum MBL level in naïve controls, spontaneously recovered individuals and nonprogressed carriers, were 1,532 ng/mL (2,571 ng/mL), 1,321 ng/mL (1,894 ng/mL), and 1,318 ng/mL (1,898 ng/mL), respectively. They were not significantly different between the three groups by Mann-Whitney U test (P > .05).
Low MBL Was Associated With Occurrence of Cirrhosis and Hepatocellular Carcinoma in Progressed Carriers.
To further study the role of mbl2 polymorphisms in the association of HBV infection, we compared the genotype distributions between nonprogressed carriers and progressed carriers (Table 3). They also were consistent with the existence of Hardy-Weinberg equilibrium. The frequencies of mbl2 genotypes were significantly different between the 2 groups with a dose-dependent effect (P = .01), that is, a higher risk for the extremely low MBL level genotype, group 3 (OR = 3.21, 95% CI: 1.43-7.21), than for the low MBL level genotype, group 2 (OR = 1.36, 95% CI: 0.76-2.43). The haplotype frequencies also were significantly different between progressed carriers and nonprogressed carriers (P = .002). The frequencies of low MBL–associated haplotypes, XA and YB, were significantly increased in progressed carriers with OR of 1.97 (1.26-3.06) and 1.90 (95% CI: 1.19-3.02), respectively.
Table 3. mbl2 Genotype and Haplotype in Nonprogressed and Progressed Carriers
Nonprogressed Carriers (n = 320)
Progressed Carriers (n = 199)
OR (95% CI)
NOTE. A dose-dependent effect of mbl2 genotype groups was shown for the risk of progression of HBV infection with OR = 1.36 for low MBL genotypes in group 2 and OR = 3.21 for extremely low MBL genotypes in group 3 (P = .01). The haplotype distribution also differed between nonprogressed carriers and progressed carriers (P = .002). The 2 low MBL level haplotypes, XA and YB, were significantly decreased in nonprogressed carriers with ORs of 1.97 and 1.90, respectively. The serum MBL level of progressed carriers was significantly lower when compared with nonprogressed carriers (P = .001).
P values and odds ratios (95% confidential interval) were calculated with the use of logistic regression models, adjusted for age, sex, ALT to ULN ratio, HBeAg status, HBV-DNA titer, and HBV genotype as covariables. The most dominant mbl2 genotype group, group 1, and haplotype, YA, were used as the reference group (OR = 1) in overall chi square analysis.
The Mann-Whitney U test was used to compare the serum MBL levels between groups.
The serum MBL level was significantly lower in progressed carriers [median (mean), 936 ng/mL (1,593 ng/mL)] than nonprogressed carriers [median (mean), 1,318 ng/ml (1,898 ng/mL)] (Mann-Whitney U tests, P = .001) (Table 3). Then, multiple linear regression was used to analyze the serum MBL levels in nonprogressed carriers and progressed carriers by adjusting mbl2 genotype group and age as covariables. No significant differences were observed (Fig. 1). The lower serum MBL concentration in the progressed carriers was attributable to the increase in the frequency of the low and extremely low MBL genotypes, groups 2 and 3.
MBL Bound to HBsAg and Mediated C4 Deposition.
We demonstrated that MBL could bind to HBsAg in a dose-dependent manner (Fig. 2). The binding was calcium-dependent and mannan-inhibitable, suggesting that MBL can recognize HBsAg through its carbohydrate recognition domain. In addition, we performed a C4 deposition assay and found an enhancement of C4 deposition on HBsAg by MBL (Fig. 3), indicating that the MBL activates complement on HBsAg binding through the lectin pathway.
We have previously identified the MBL codon 54 allele B as a risk factor in HBV-related cirrhosis.27 Although lower serum MBL levels in HBsAg carriers were also observed in the previous study, this could not be entirely explained by increases in the frequency of the B allele.27 We have demonstrated in the current study that, in addition to allele B, the promoter polymorphism (−221 X/Y) is associated with the occurrence of the cirrhosis or hepatocellular carcinoma in progressed carriers after controlling for age, sex, ALT ratio to ULN, HBeAg status, HBV DNA titer, and HBV genotype. We further showed that the lower MBL level observed in the progressed carriers when compared with nonprogressed carriers can be explained entirely by the overrepresentation of the low MBL genotypes in the progressed carriers (Fig. 1). This indicates that the low MBL genotypes revealed in this study are fully responsible for the lower MBL level in progressed HBsAg carriers. To elucidate the mechanism of MBL action in HBV infection, we performed in vitro assays to show that MBL could bind to HBsAg and could, in turn, mediate C4 deposition. This implied that MBL could be involved in the clearance or control of HBV. Our findings further support an important role for MBL as an innate molecule in host defense against infectious diseases.15–19
Thomas et al.29 reported an increased frequency of the allele D (codon 52) of mbl2 in Caucasian but not Asian patients with chronic HBV infection.29 However, in the Chinese allele C and allele D of mbl2 are extremely rare, if not entirely absent.22, 23, 27 Another report suggested that allele B (codon 54) was over-represented in Vietnamese patients with acute hepatitis B when compared with healthy controls.35 However, we did not find any increase in low MBL genotype frequency in spontaneously recovered individuals or nonprogressed HBsAg carriers compared with naïve controls to implicate low MBL levels as a susceptibility factor in acute HBV infection. In the Vietnamese study, the number of patients with acute HBV infection was relatively small at 31, and the healthy controls, which were defined as HBsAg negative, may have included both naïve and spontaneously recovered individuals. In contrast, two negative reports have come from Gambia and Germany using patient cohorts of 180 and 89, respectively.30, 31 In this study, the association between low MBL genotypes and the occurrence of cirrhosis and hepatocellular carcinoma in HBsAg carriers was documented by studying over 1,000 Chinese. To further confirm this association in other populations, replicate study with enough sample size should be performed in different ethnic groups.
The mean age of our nonprogressed carriers was lower than that of the progressed carriers and, hence, we have controlled for the effect of age and sex in all analyses by logistic regression. Moreover, we have shown in our previous study that mbl2 genotype/haplotype frequencies were not significantly different among different age groups in healthy controls.32 Interestingly, for our nonprogressed carriers, there was a suggestion of an increase in group 1 genotype frequency and a decrease in the groups 2 and 3 genotype frequencies in the 45- to 69-year age group when compared with the 14- to 44-year age group (Table 4). This outcome suggests that, with aging, individuals in the mbl2 genotype group 1 would be more likely to remain nonprogressed than those in groups 2 and 3. Twenty-five percent of HBsAg carriers will progress and develop cirrhosis or hepatocellular carcinoma.36 Hence, the decrease in the frequencies of genotype groups 2 and 3 in the 45- to 69-year age group of nonprogressed carriers is likely to be attributable to this progression. This is in line with our observation that low MBL genotypes are associated with the occurrence of cirrhosis and hepatocellular carcinoma in progressed carriers. Also, the lower percentage of HBeAg seropositivity in progressed carriers (Table 1) was probably attributable to the effect of age, because HBeAg seroconversion would occur with aging.37 There was no difference in ALT ratio to ULN, HBV genotype distribution between nonprogressed and progressed carriers, whereas there was a marginal difference for HBV DNA titer.
Table 4. Genotype and Haplotype Frequencies of mbl2 Polymorphisms in Nonprogressed Carriers by Age Group
Non-progressed Carriers (n = 320)
Age group in years
14-44 (n = 251)
45-69 (n = 69)
NOTE. An increase in frequencies in group 1 and a decrease in frequencies in groups 2 and 3 were observed in the 45- to 69-year-old age group compared with the 14- to 44-year age group, although statistical significance was not reached.
Mean age (y)
32.3 ± 8.0
51.6 ± 6.6
YA/YA and YA/XA
YA/YB and XA/XA
YB/XA and YB/YB
Two possible mechanisms may explain how low MBL levels may lead to progression of HBV infections. First, MBL may have a direct effect on HBV infection through complement activation. Because a mannose-rich oligosaccharide is found on the preS2 region of the middle hepatitis B surface protein,28 MBL could theoretically bind HBV. To substantiate this hypothesis, we have studied the binding of MBL to HBsAg. As shown in Fig. 2, a calcium-dependent MBL binding to HBsAg was observed, and this binding was substantially inhibited by mannan. These data provide evidence that MBL is able to bind to HBsAg via its multiple carbohydrate recognition domains and suggest that MBL may function as an opsonin for HBV. Also, MBL plays an important role in the complement system by acting as the recognition molecule of the lectin pathway.12 MBL binds to MASP2, and the latter is responsible for the activation of C4 and C2, thereby generating C3 convertase, C4b2a, which in turn converts C3 to C3b. To study the possible role of MBL on complement activation by HBsAg, we incubated HBsAg with MBL and C4 in vitro. If there is complement activation, C4 should be cleaved by MASPs, and the resulting C4b fragments will form covalent bonds with nearby hydroxyl or amino groups. C4 deposition was indeed observed and was MBL dose dependent in our experiment (Fig. 3). This indicated that MBL activates complement on HBsAg-MBL complexes through the lectin complement pathway and therefore may be involved in HBV clearance.
Apart from the possibility of direct HBV clearance, MBL-mediated complement activation could be involved in immune complex removal during HBV infection. Increased levels of immune complexes in chronic liver disease have been reported and are thought to be associated with the inflammatory responses in liver damage.38, 39 Complement receptor type 1 (CR1) on circulating blood cells, including erythrocytes and monocytes/macrophages, can bind C3b, C4b, and possibly MBL.40, 41 This interaction allows circulating cells to direct immune complexes to the organs of the reticuloendothelial system for safe removal. Also, Roos et al.42 demonstrated that the binding of MBL to immunoglobulin A can lead to complement activation, and MBL has been shown to be involved in removal of different immune complexes.43 Therefore, low MBL levels may lead to defective complement activation, poor clearance of immune complexes with subsequent deposition in the liver, and inflammation.
A second possible mechanism is through the regulation of inflammatory cytokines by MBL. High MBL levels have been shown to be involved in decreasing the production of inflammatory cytokines, such as IL-6, IL-1β, and TNF-α, by monocytes in response to meningococci, whereas low MBL concentrations can enhance the production of IL-6 and IL-1β.20 The pathogenesis of cirrhosis and hepatocellular carcinoma is a multistage process involving chronic liver cell injury, inflammation, and hepatocyte regeneration. The low MBL levels, which associate with high production of inflammatory cytokines such as IL-6,20 may promote chronic inflammation and oxidative stress. Enhanced production of IL-6, a fibrogenic cytokine, can activate stellate cells to contribute to fibrogenesis and increase proliferation of hepatocytes for further persistent HBV infection, which leads in turn to cirrhosis and hepatocellular carcinoma.44, 45 Chronic inflammation has been shown to mediate liver cell injury and the development of hepatocellular carcinoma in the absence of viral transactivation, insertional mutagenesis, and genotoxic chemicals.46 The oxidative stress coupled with increased accumulation of reactive oxygen species, DNA damage, changes in cellular proliferation and, apoptosis or necrosis in hepatocytes could contribute to the development of hepatocellular carcinoma.44, 47
We have shown that serum MBL levels and mbl2 polymorphisms were similar to control values in spontaneously recovered individuals and nonprogressed carriers. However, we cannot entirely exclude a role for MBL in acute HBV infection. Cell-mediated immune responses have been considered to be the main mediators of HBV clearance and cytotoxic T cells are involved in acute progressed HBV infection in humans.48 However, using an HBV transgenic mouse model and acutely infected chimpanzees, only a minority of infected hepatocytes were eliminated by the direct action of cytotoxic T cells.49, 50 The antiviral function was mediated by interferon (IFN)-γ and TNF-α secreted by the cytotoxic T cells or by the antigen-nonspecific macrophages.49, 50 Therefore, MBL also may be involved in the antiviral response by regulating cytokines such as TNF-α.20
In conclusion, the mbl2 polymorphisms that result in low serum MBL levels associate with the occurrence of cirrhosis and hepatocellular carcinoma in progressed carriers. Moreover, MBL is able to bind to HBsAg, resulting in complement activation. These findings should provide new approaches to elucidating the mechanisms of clearance of HBV and HBV containing immune complexes as well as understanding disease progression.
We thank Dr. Olaf Neth and Ms. Clare Booth for preparations of plasma-derived MBL. We would also like to thank Prof MW Turner for stimulating discussion and comments.