Is hepatitis B antigen in cord blood an immunotolerogen playing a critical role in the pathogenesis of chronic hepatitis B?

ABSTRACT

Recently, we showed that infants with or without fetal hepatitis B e antigen (HBeAg) exposure had comparable antibody response to hepatitis B vaccination and proposed that fetal HBeAg exposure appears to not induce immunotolerance to HBV. Here, we summarized the different results on the topic of fetal HBeAg exposure in inducing immunotolerance to HBV nucleocapsid protein, and the evidence to back up that the tendency of high infection rate in infants of HBeAg-positive mothers is more likely associated with higher maternal viral loads and is less likely associated with the fetal HBeAg exposure. We consider that whether fetal HBeAg exposure plays an important role in the pathogenesis of chronic HBV infection remains to be an open question.

KEYWORDS:

• HBeAg
• fetal exposure
• chronic HBV infection

We appreciate Dr. Milich’s comments on our article recently published in Human Vaccine & Immunotherapeutics,¹ in which we concluded that the transplacentally acquired maternal hepatitis B e antigen (HBeAg) in utero appears not inducing immunotolerance to hepatitis B virus (HBV) and may be not associated with the pathogenesis of chronic HBV infection after neonatal exposure to HBV, based on the findings of similar levels of antibody response to hepatitis B vaccination between infants with and without fetal exposure to maternal HBeAg and between infants with relatively high and low fetal HBeAg exposure respectively.² Dr. Milich considered that our conclusion is not supported by our findings, because we did not directly measure the immune tolerance to HBeAg in HBeAg-exposed and -unexposed infants and did not determine the chronic HBV infection rate in infants born to HBeAg-positive and -negative mothers in the absence of immunoprophylaxis.¹

Indeed, we did not measure the T cell responses to HBV antigens, such as hepatitis B surface antigen (HBsAg), HBeAg, and hepatitis B core antigen (HBcAg) in our study, and we could not compare the specific T cell responses in HBeAg-exposed and -unexposed infants. However, previous studies demonstrated that production of antibody against HBsAg (anti-HBs) is dependent on the participation of T cells.^3,4 The comparable robust anti-HBs response to hepatitis B vaccine (recombinant HBsAg) in infants with or without fetal HBeAg exposure indicates that the T cell response to HBsAg were not inhibited by the transplacentally acquired maternal HBeAg in fetus, implying that fetal HBeAg exposure may not induce immunotolerance to HBsAg. Considering that HBsAg constitutes the major protein on the envelope of HBV, it is reasonable to deduce that no immunotolerance to HBsAg means no immunotolerance to HBV. Thus, we consider that the findings in our article provided, at least to some extent, evidence to support the conclusion that “HBeAg appears not to be inducing immunotolerance to HBV.”

Dr. Milich mentioned that we did not determine the rate of chronic HBV infection in neonates born to HBeAg-positive and -negative mothers in the absence of immunoprophylaxis.¹ Now it is not ethical for anyone to determine the infection rate in the absence of immunoprophylaxis, because administration of hepatitis B immunoglobulin (HBIG) and/or hepatitis B vaccine in neonates of HBV carrier mothers is a standard protocol to prevent perinatal infection of HBV. A large amount of historical data showed that the chronic infection rate in children born to HBV carrier mothers with positive or negative HBeAg is 70–90% and 10–30%, respectively, before the availability of immunoprophylaxis.^5–7 Here, we would like to point out that the relatively lower rate (10–30%) of chronic HBV infection in infants born to HBeAg-negative carrier mothers does not mean that 70–90% other infants without chronic infection had underwent acute infection; only a small proportion of them had experienced acute infection, with the presence of anti-HBs and/or antibody against HBcAg (anti-HBc), and the majority of them had no evidence of infection because of the absence of both anti-HBs and anti-HBc: they had just exposed to, but not infected with HBV.⁸

Based on the substantial difference in the chronic HBV infection rate in infants born to HBeAg-positive and -negative mothers in the absence of immunoprophylaxis, Dr. Milich’s group hypothesized HBeAg as an immunoregulatory protein. Using a mouse model, Dr. Milich’s group found that transgenic mice expressing HBeAg in the circulation were T-cell tolerance to both HBeAg and HBcAg, and the nontransgenic offsprings of the HBeAg-expressing mice showed reduced T-cell responsiveness to HBeAg.⁹ Based on these data, they proposed that HBeAg possesses an immunoregulatory function in utero by establishing T-cell tolerance to HBeAg and HBcAg, which may make neonates of HBV-infected mothers to be persistently infected.⁹ However, it is unknown whether the transgenic mice expressed HBcAg in the liver because the article did not present relevant data.⁹ Since the transgenic gene in this mouse model covered the encoding sequence for HBcAg, which is not secretory in mammalian cells or Xenopus oocytes,^10,11 the expression of HBcAg in liver should be checked. The hypothesis of the cross T-cell tolerance to HBcAg induced by HBeAg should be established after exclusion of HBcAg expression in the liver. Moreover, it was unknown whether the nontransgenic offsprings of the HBeAg-expressing mice had reduced T-cell responsiveness to HBcAg.⁹

Dr. Milich cited a reference (ref 8 in the Letter) in that Hsu et al. reported that cord blood mononuclear cells from neonates born to healthy (normal control), HBeAg-negative carrier (no fetal HBeAg exposure), and HBeAg-positive (fetal HBeAg exposure) mothers did not proliferate to recombinant HBcAg at all.¹² Dr. Milich agreed on Hsu et al’s explanation that the unresponsiveness was resulted from immune tolerance to HBeAg/HBcAg at the T-cell level, because of HBeAg exposure in utero.¹ However, we think that this conclusion was somewhat far-fetched, because those healthy and HBeAg-unexposed neonates also had no responsiveness to HBcAg. It was likely that the unresponsiveness in HBeAg-exposed neonates was associated with the cord blood T cells not primed by HBcAg in utero, as in utero HBV transmission very rarely occurs.⁸ The presence of HBeAg in cord blood meant that mononuclear cells had been primed by HBeAg, but the response to HBeAg was not tested in Hsu et al.’s study,¹² which makes the interpretation of results more complicated. Therefore, we do not consider that Hsu et al.’s study added conceivable evidence to support the T cell immune tolerance to HBcAg by fetal exposure to maternal HBeAg.

On the other hand, other scholars obtained the opposite data based on a similar transgenic mouse model expressing HBeAg and considered that HBeAg does not tolerize cytotoxic T cells.^13,14 Furthermore, studies in infants born to HBV-infected mothers with different HBeAg status demonstrated that HBV exposure in utero induces a state of trained immunity characterized with enhanced innate immune cell maturation and increased Th1 development, rather than immunotolerance.¹⁵ Therefore, whether in utero exposure to maternal HBeAg can really induce T-cell immunotolerance to HBcAg requires further investigation.

We understand that the assumed immunotolerance induced by HBeAg exposure in utero refers to T-cell specific to HBeAg and cross-specific to HBcAg, which has been used to explain the high chronic infection rate in infants born to HBeAg-positive mothers.⁹ We did not state that non-immunotolerance to HBsAg means non-immunotolerance in HBe/HBcAg-specific T-cell response, and just stated that non-immunotolerance to HBsAg may indicate non-immunotolerance to HBV in our article.² In his letter, Dr. Milich emphasizes the role of HBeAg in promoting chronicity.¹ However, with the development of quantification assay of HBV DNA, it has been demonstrated that the viral loads in HBeAg-positive pregnant women are usually 1000–10,000 times higher than those in HBeAg-negative pregnant women.^16–18 This can explain the much higher chronic infection rate in infants born to HBeAg-positive mothers before the availability of immunoprophylaxis and the still-occurring immunoprophylaxis failure in infants born to HBeAg-positive mothers but almost no immunoprophylaxis failure in infants born to HBeAg-negative carrier mothers.^8,19–21 Additionally, antiviral therapy during the third trimester in HBeAg-positive pregnant women, which significantly reduces maternal HBV DNA levels around delivery but rarely reduces maternal HBeAg levels,²² can almost block the mother-to-infant transmission of HBV.^22–24 These results reinforce the role of high viral titers, rather than the maternal HBeAg itself, in the pathogenesis of chronic infection in infants born to HBeAg-positive mothers. Since these contents are not closely associated with our study on the comparison of anti-HBs levels in infants with or without fetal HBeAg exposure, we did not comment on these issues in our published article.² Moreover, combined use of HBIG and hepatitis B vaccine can prevent 90–95% of chronic infections in infants born to HBeAg-positive mothers,^8,19–21 and timely use of them can prevent 97–98% of the chronic infection,^22,24,25 in which the scenario of maternal HBeAg exposure in fetuses and neonates is not altered and the development of anti-HBs response should play a critical role. Thus, we considered that the potent anti-HBs response in our infants provided evidence from another aspect to support the hypothesis that the transplacentally acquired maternal HBeAg in utero may be not associated with the pathogenesis of chronic HBV infection after neonatal exposure to HBV.²

Additionally, although it is usually considered that the specific T cell responses to HBcAg play critical roles in eliminating HBV infection in hepatocytes,²⁶ the failure of therapeutic vaccine containing HBcAg cytotoxic T lymphocyte epitope (amino acids 18–27) in the treatment of chronic hepatitis B,²⁷ suggests that the mechanisms for the chronicity is much more complex beyond our imaginations.

Dr. Milich considered that our statement that “the role of HBeAg in inducing neonatal immunologic tolerance to HBV remains controversial” is ambiguous,¹ however, we think that it has no problem, since different studies on this topic presented contrast results as explained above.^9,13,14 In the Filed Virology (4th ed), Drs Holliger and Liang also stated that “…. Although this is an attractive hypothesis, the role of HBeAg as a neonatal immunologic toleragen remains controversial”.²⁸

In summary, the similar anti-HBs response to hepatitis B vaccination in infants with or without fetal HBeAg exposure demonstrated that fetal HBeAg exposure does not cause immune tolerance to HBsAg, suggesting that HBeAg appears not to induce immune tolerance to HBV since HBsAg is the major protein on the HBV envelope. Whether fetal HBeAg exposure plays a critical role in the pathogenesis of chronic HBV infection remains to be an open question.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.