Hepatitis B virus lymphotropism: emerging details and challenges

Abstract

The hepatitis B virus (HBV) is predominantly a hepatotropic virus but also infects cells of the lymphatic system. HBV genomes (DNA, messenger (m)RNA, covalently closed circular (ccc) DNA) and proteins have been found in extrahepatic sites such as peripheral blood mononuclear cells (PBMC), lymph nodes, spleen, bone marrow and cerebrospinal fluid. HBV entry into hepatocytes occurs by binding of the HBV preS1 surface protein to its specific receptor, the bile acid transporter, sodium taurocholate co-transporting polypeptide (NTCP). Although the mechanism of HBV entry into lymphatic cells is unknown, the pre S1 encoded surface protein is thought to be involved. Extrahepatic HBV infection has been studied in both chronic HBV (CHB) and in occult HBV infection (OBI). Studies have shown that HBV genomes are present in different PBMC subsets from chronically infected carriers. Unique HBV variants have been found in PBMC compared to plasma or liver in both nucleos(t)ide analogue (NA) treated and untreated CHB carriers, suggesting replication and compartment specific evolution of HBV. In HBV coinfection, HBV genomes were found in PBMC from hepatitis C virus (HCV), human immunodeficiency virus (HIV) and hepatitis delta virus (HDV) co-infected individuals. Moreover, during pregnancy, the trans placental passage of HBV infected PBMC from highly viremic mothers to infants is one of the postulated means of vertical transmission of HBV. Taken together, HBV infection in extrahepatic sites (i.e., PBMC) is implicated in multiple facets of HBV pathogenesis such as persistence, viral evolution and vertical transmission.

Keywords:

• Hepatitis B virus persistence
• peripheral blood mononuclear cells
• compartmentalization
• occult HBV infection

Introduction

The hepatitis B virus (HBV) is a global human pathogen that causes chronic hepatitis B (CHB) in ~250 million people worldwide who have a 15–40% lifetime risk of end-stage liver disease and/or hepatocellular carcinoma (HCC)(WHO Hepatitis B factsheet, 2017). The HBV is a small hepatotropic DNA virus of the Hepadnaviridae family. Hepadnaviruses are found in mammals such as woodchuck, humans, ground squirrel (genus: orthohepadnavirus) and birds such as duck and goose species (genus: avihepadnavirus) (Schaefer, 2007; Seeger & Mason, 2015). Hallmarks of the hepadnavirus replication cycle include: (1) formation of intranuclear minichromosome, known as covalently closed circular DNA (cccDNA) and (2) the reverse transcription of a pre-genomic RNA (pgRNA) in core particles used for synthesis of the relaxed circular DNA (rcDNA) genome (Hu & Seeger, 2015). This cccDNA molecule is the key viral persistence marker in infected cells and serves as a reservoir for the HBV (Nassal, 2015). Due to the error-prone reverse transcription of an RNA intermediate to copy its genome, HBV exists as a pool of viral variants or quasispecies. There are 10 HBV genotypes reported (A-J) with specific geographical distribution (Gao, Duan, & Coffin, 2015). There is evidence that the HBV genotype, viral diversity and mutants correlate with the natural history of chronic hepatitis (Li, Wang, Li, & Ding, 2017; Locarnini & Yuen, 2010) .

HBV tissue tropism

Although HBV is considered a hepatotropic virus, it can also infect and propagate within cells of the immune system (i.e., lymphoid cells), which is an important immune evasion strategy (Coffin et al., 2015; Gao et al., 2017). The pre-S1 encoded surface protein is involved in virus–host interaction and virus entry into cells (Pontisso et al., 1991; Pontisso & Alberti, 1991). In hepatocytes, the sodium taurocholate co-transporting polypeptide (NTCP) receptor facilitates virus entry (Tong & Li, 2014). However, the exact mechanism for entry into PBMC has not yet been identified. HBV replicative intermediates and proteins have been detected in extrahepatic sites such as spleen, lymph nodes, bone marrow and peripheral blood mononuclear cells (PBMC)(Coffin et al., 2011a; Lu et al., 2009; Umeda et al., 2005). Zeldis et al. found that HBV can infect immature hematopoietic stem cells: myeloblasts, normoblasts and lymphoblasts. Hepatitis B surface antigen (HBsAg) was detected in nuclei of 70% of the infected cells by electron microscopy (Zeldis, Mugishima, Steinberg, Nir, & Gale, 1986). In a study, it was noted that 28/32 (87%) of CHB carriers had at least one HBV marker (Hepatitis B e antigen, HBeAg, HBsAg and/or HBV DNA) detected in PBMC (Bouffard, Lamelin, Zoulim, Pichoud, & Trepo, 1990). Extrahepatic persistence of HBV in lymph nodes, spleen, bone marrow, kidney, skin, colon, stomach, testes and peri-adrenal ganglia was demonstrated in 3/4 patients with HBV related end-stage liver disease (Mason, Wick, White, & Perrillo, 1993). An in vitro study showed that HBV is selectively able to infect immature myeloid cells vs. differentiated mature cells (Sing, Prior, Fernan, & Cooksley, 1993). In all studies, the liver is the main reservoir for the HBV but low-level HBV infection is ongoing within extrahepatic sites. In occult HBV infection (OBI), individuals are serum HBsAg negative and have natural immunity (with HBV core and/ or HBsAg antibodies) and HBV DNA remains detectable in liver/serum/PBMC (Raimondo, Pollicino, Romanò, & Zanetti, 2010). Interestingly, low level HBV infection, including OBI appears to be predisposed to infect cells of the lymphatic system (Coffin et al., 2011a, 2011b).

Optimal methods for confirming HBV lymphotropism

Confirmation of a productive intracellular (hepatic or extrahepatic) infection requires detection of HBV replicative intermediates such as total DNA, mRNA and cccDNA, as well as viral proteins. Intrahepatic viral genome levels are often correlated with levels of serum HBV-DNA, serum quantitative surface antigen (qHBsAg), pre-genomic (pg) RNA and HBV core related antigen (HBcrAg). Older studies relied on less sensitive nucleic acid hybridization assays for detection of HBV species (i.e. HBV cccDNA, relaxed-circular DNA and mRNA), with detection limit of 10⁵ copies per mL in liver tissue (Samuel & Roque-Afonso, 2007; Liu & Yao, 2015). Due to much lower level HBV infection in PBMC, the detection of HBV DNA and/or cccDNA by a direct nucleic acid hybridization is often difficult. Therefore, current molecular assays that are polymerase chain reaction (PCR) based are widely used in detection of low levels of HBV DNA, especially within PBMC. The detection limits of the most sensitive nested PCR/nucleic acid hybridization (NAH) assays available are below 10 virus genome copies or equivalents (vge) per mL (Datta, Chatterjee, & Veer, 2014; Mulrooney-Cousins & Michalak, 2007). Due to high sensitivity of PCR, the issue of cross contamination during DNA extractions as well as possibility of extracellular adhered viral particles must be addressed to avoid spurious detection. Thus, established and rigorous protocols for low-level extrahepatic HBV detection (i.e. in PBMC) should include: (i) standard PCR and mock extraction controls; (ii) extracellular enzymatic treatment to removed adhered virus using DNase and trypsin treatment. Full genome sequencing is ideal but challenging for detection of low viral loads. The protein free HIRT DNA extraction is the most widely used protocol for cccDNA isolation (Cai et al., 2013). Both protein free relaxed circular DNA as well as cccDNA can be extracted simultaneously using this method. T5 exonuclease is thus far the best enzyme to digest out RC-DNA leaving behind cccDNA. However, there is controversy about the optimal duration of enzymatic digestion (Schreiner & Nassal, 2017).

Experimental data supporting hepadnaviral lymphotropism

The Woodchuck hepatitis virus (WHV) model of HBV infection

WHV and HBV belong to the family of Hepadnaviruses. Multiple studies in the WHV model have supported the lymphotropic nature of hepadnaviral infection. WHV lymphotropism is an intrinsic property of WHV rather than cell type specific infectivity of viral variants (Mulrooney-Cousins & Michalak, 2008). In woodchucks infected with low doses of virus (<300 copies/mL) the WHV exclusively infects lymphoid cells (Michalak, Mulrooney, & Coffin, 2004). Offspring born to woodchuck dams with occult WHV infection following recovery from acute viral hepatitis, develop occult WHV infection that was restricted to the lymphoid system. Both plasma and lymphoid derived virus from these woodchuck offspring induced classical acute WHV hepatitis in virus-naïve woodchucks (Coffin & Michalak, 1999). Additionally, WHV genomes were detectable by in situ PCR combined with flow cytometry in 3–20% of infected cells from woodchucks with symptomatic WHV hepatitis (Mulrooney & Michalak, 2003). WHV replication in PBMC is upregulated following ex-vivo mitogen stimulation and virus replicating in woodchuck lymphoid cells is infectious to hepatocytes and lymphoid cells in vitro (Coffin & Michalak, 1999). Integrated WHV DNA was detected in both liver and lymphoid cells in woodchucks at 2.5–3.5 years post primary occult WHV infection (Mulrooney-Cousins, Chauhan, Churchill, & Michalak, 2014) and it has been recently shown that the onset of hepadnaviral induced oncogenesis can begin as early as 1 h post-infection in woodchucks (Chauhan, Churchill, Mulrooney-Cousins, & Michalak, 2017). Considering the significant virological similarities & patterns of disease progression exist between HBV and WHV, it is conceivable that these findings might also be applicable to human HBV pathogenesis. Extrahepatic tissue tropism of HBV is less well studied due to accessibility and feasibility of obtaining lymphoid tissues, however several studies have investigated HBV infection in PBMC.

Compartmental evolution of HBV

Studies from our group have noted diverse HBV genotypes in plasma vs. PBMC compartments in treatment naïve patients, including unique sequences in plasma and PBMC. Immune escape mutations were more frequent in PBMC of immune-active patients as compared to other CHB disease phases, while drug resistant and immune escape mutants were found in plasma vs PBMC. These data suggest disease phase as well as compartment specific evolution of HBV (Coffin et al., 2015). HBV within PBMC is transcriptionally active with replicative intermediates and variants despite long term NA therapy (Coffin et al., 2011a; Mason et al., 1993). We recently found that HBV genotypes differ at baseline and after NA therapy (~1 year post therapy) between different compartments-liver, plasma and PBMC. Importantly a decline of 200 copies/mL HBV DNA in plasma occurred post treatment but no significant decline of intracellular HBV DNA and cccDNA was noted. Another interesting observation was that genotype switch occurred in PBMC and plasma compartments before and after therapy, while this phenomenon did not occur in the liver (Gao et al., 2017). A study from India showed unique sequences in PBMC vs. plasma in CHB carriers. Clonal sequencing showed that genotype A virus (sub-genotype Ae/A2) with G145R vaccine escape mutation was exclusively present in PBMC but not serum (which had sub-genotype A1) (Datta et al., 2009). The emergence of compartment specific variants has been attributed to genotype specific differences in pre-genomic RNA secondary structures (Datta & Chakravarty, 2016; Kidd-Ljunggren, Zuker, Hofacker, & Kidd, 2000).

The clinical relevance of HBV lymphotropism

The significance of HBV lymphotropism is controversial, yet studies by our group and others suggest an important role in the following: (1) viral persistence, especially occult HBV in HBV mono-infected and in HBV/human immunodeficiency virus (HIV) or hepatitis C virus (HCV) co-infected individuals; (2) persistence after liver transplantation; and (3) vertical transmission of HBV.

Viral persistence after liver transplantation

Research by our group in patients on long-term anti-HBV therapy has shown that the PBMC reservoir is more likely to carry the wild-type (WT) virus that might be related to the lower level of antiviral therapy penetration into lymphoid cells and the limitations of the available replication space (Coffin et al., 2011b).

HBV carried within PBMC has been reported to be associated with re-infection in liver transplant (LT) recipients (Brind et al., 1997; Roche et al., 2003). The recurrence of HBV post-LT can be demonstrated first in PBMC followed by liver, and the predominant HBV surface (S) gene sequence in PBMC before LT is the dominant sequence identified after LT (Féray et al., 1990). Our prior studies found that the majority of LT recipients on effective prophylactic therapy had detectable HBV-DNA in PBMC and/or liver suggesting an ‘occult HBV infection’ (Coffin et al., 2011b). However, the potential impact of extrahepatic (i.e. PBMC) HBV on relapse risk following withdrawal of long-term prophylaxis is not clearly understood (Raimondo et al., 2010).

Viral persistence in occult hepatitis B infection (OBI) and co-infections

In OBI, despite HBsAg loss and natural immunity HBV-DNA remains detectable at low levels (<10³ copies/mL) in liver, serum and PBMC (Raimondo et al., 2010). Occult HBV persistence might occur as a result of resolution of acute hepatitis or following an asymptomatic viral encounter (Makvandi, 2016). Due to shared routes of transmission, co-infection of HBV with HCV/HIV/HDV as well as OBI in HCV and/or HIV infection is common. OBI is frequently recognized in chronic hepatitis C infections. OBI in chronic HCV can accelerate liver disease (Cacciola et al., 1999; Giannini et al., 2003) and in HIV/HCV may lead to more rapid progression to acquired immunodeficiency syndrome (Wang et al., 2012). In a study, using sensitive nested PCR, HBV-DNA was detected in 20% of plasma samples and 32.6% PBMC (Vakili Ghartavol et al., 2013). Sagnelli et al. found that 50% of the HCV patients with detectable HBV-DNA in PBMCs had negative results in plasma (Sagnelli et al., 2001). In another study, 119 HBsAg negative, chronic hepatitis C patients were screened for the presence of OBI. 40% of the patients had prior HBV infection and had severe liver disease compared to those without prior exposure (Giannini et al., 2003).

Our prior studies on OBI in HIV+, anti-HBc+ patients on anti-HBV/HIV therapy in a cohort of patients followed long term demonstrated 42% (19/45) HBV-DNA positivity especially in PBMC (18/45), including some cases positive for cccDNA. This study suggests HBV lymphotropism in HIV+ patients, as HBV genomes were detected in PBMC and not in plasma for most cases. The major HBV genotypes identified within PBMC was C or D (Coffin et al., 2014). HBV genomes have also been found in PBMC from HBV/HCV, HBV/HIV and HBV/HDV co-infection cases. In a cohort of HBV/HCV co-infected patients, HBV-DNA and HCV-RNA were detected in PBMC in 10% of cases, suggesting that both viruses can co-exist within PBMC (Coppola et al., 2008). However, we noted profound suppression of HCV by HBV superinfection in PBMC in a patient with HBV/HCV co-infection. HBV DNA and cccDNA were detected in PBMC indicating active replication but HCV-RNA remained undetectable in the same cells (Coffin, Mulrooney-Cousins, Lee, Michalak, & Swain, 2007). Longitudinal studies in a larger cohort of HBV/HCV co-infected patients is necessary in order to understand dynamics of HBV within PBMC in co-infection settings. A study by Magrin et al., reported the presence of HBV-DNA within PBMC in 8/23 HBV/HDV co-infected individuals (Magrin et al., 1989). Another report showed that in 30 co-infected samples demonstrated by nucleic acid hybridization assay (Southern blot) that PBMC HBV DNA was detectable in ~45% of the samples (Moraleda, Bartolomé, Martinez, Porres, & Carreño, 1990). These observations show that HDV does not affect the presence of HBV within PBMC. However, more studies are necessary to confirm these findings, including functional studies of PBMC-derived virus. Several older studies in HBV/HIV individuals, have confirmed HBV-DNA in PBMC by dot blot and southern blot hybridization (Bartolom et al., 1990; Laure et al., 1985). Our study compared detection of HBV genomes in PBMC subsets – CD4+ T, CD8+ T, CD14+ monocytes, CD19+ B, CD56+ NK cells in 14 HBV mono-infected and 6 HBV/HIV co-infected cases. HBV replicative intermediates could be detected within all the PBMC subsets consistent with other studies (Chemin et al., 1992; Trippler, Meyer zum Büschenfelde, & Gerken, 1999; Yoffe, Noonan, Melnick, & Hollinger, 1986). Interestingly, in the co-infected cohort, HBV was not detectable in CD4+ T cells. Overall, these findings suggest PBMC as an extrahepatic HBV reservoir and also that HIV coinfection impacts HBV lymphotropism (Lee et al., 2015).

Role of HBV lymphotropism and vertical as well as horizontal transmission

HBV vertical or mother-to-child transmission (MTCT) usually occurs at the time of birth / perinatal period due to contact with infected maternal body fluids, however other postulated mechanisms of MTCT include passage of HBV virions or infected peripheral blood mononuclear cells (PBMCs) through the placenta (Bai, Li, Yue, & Shi, 2011; Xu et al., 2015). Moreover, successfully vaccinated uninfected infants born to HBsAg+/HBeAg− mothers carriers exhibit HBV specific T helper cell responses, suggesting in utero encounter to viral antigens (Koumbi et al., 2010). Bai et al. reported 16/60 (26.7%) infants born to mothers showing HBV-DNA positivity in PBMC became infected with HBV (Bai et al., 2011). In another study, 81/142 infants born to mothers with HBV-DNA in PBMC acquired HBV infection. In contrast, a study by Poovorawan et al., found that PBMC from neonates born to CHB mothers were not HBV infected despite >70% positivity in mothers’ PBMC (Poovorawan, Chongsrisawat, Theamboonlers, Vimolkej, & Yano, 1997). A study showed that there was no co-relation between placental HBV infection and intrauterine infection (Shao, Zhao, & Yao Li, 2013). Our unpublished data suggests that the presence of basal core promoter (BCP) and/or pre-core (C) mutants in PBMC of mothers did not cause an overt HBV infection in infants that received complete immunoprophylaxis regimen (Joshi et al., International HBV molecular biology meeting 2017 abstract, unpublished). Collectively, these data indicate a role of PBMC in vertical transmission of HBV, however functional characterization of PBMC carrying HBV is needed. In addition, a long-term follow-up of infants that acquired intrauterine HBV infection is essential. Intrafamilial horizontal transmission of the G145R mutant found in serum of vaccinated family members has been demonstrated (Oon, Chen, Goo, & Goh, 2000). Interestingly, a specific PBMC restricted HBV subtype adw G145R mutant was identified among family members indicating probable transmission via non-sexual routes. This vaccine escape mutant was circulating within the members although they were unvaccinated. A study from India reported the occurrence of similar HBV sequences in PBMC from members of the same family suggesting horizontal transmission of HBV (Chakravarty, Neogi, Roychowdhury, & Panda, 2002).

Controversies and future directions

The PMBC is the most widely studied extra hepatic site owing to easy accessibility, but it is also important to study other lymphoid organs if possible such as spleen, lymph nodes and bone marrow. Animal models of HBV infection (non-human primates – chimpanzees and tupaias) are available. However, studies in these models are also limited due to ethical issues, fewer numbers of sanctioned animals and expertise. Studies in humanized transgenic mice susceptible to HBV infection can serve as a powerful tool in studying the complete HBV replication cycle in liver, but are not ideal to evaluate extrahepatic HBV persistence since they show sub-optimal T and B cell responses, and the viral replication occurs at a much slower rate compared to humans. Currently, the most feasible option is in vitro studies using hepatocyte cell lines.

Although there is evidence to suggest that HBV evolves differently in PBMC vs. plasma, there have been no studies reported that prove the infectious capacity of virus derived from PBMC. These could be accomplished using infection studies in cell culture models such as HepaRG cells, HepG2 cells transfected with NTCP receptor and humanized mice hepatocytes. The presence of HBV-DNA within macrophages, monocytes and B cells require careful assessment since these cells act as scavengers and can take up the virus passively. Thus, HBV functional studies in these subsets as well as within T cells, including identification of the exact receptor on each cell that aids virus entry into cells are necessary. Other relevant questions include determining the cell type specificity of HBV genotypes and variants within PBMC, especially in special populations such as pregnant CHB carriers and in co-infected patients. Ultimately, finding an effective anti-viral that would be able to infiltrate into PBMC and target HBV cccDNA may be needed to achieve a virological cure for chronic HBV infection. Addressing these questions might help in framing recommendations on diagnostic testing using whole blood or PBMC especially in blood and organ donors.

Key points

	HBV is primarily a hepatotropic virus but infects extrahepatic sites such as cells of lymphoid and myeloid origin. However, due to ease of accessibility, PBMC is the most widely studied extrahepatic site to study HBV lymphotropism.
	Evidence suggests the role of HBV carried within PBMC mainly in (a) occult HBV in mono and co-infected cases, (b) viral persistence post liver transplantation and (c) in vertical transmission.

• It is imperative to conduct functional studies confirming replication and transmissibility of PBMC derived HBV to establish clinical relevance of HBV lymphotropism. Additionally, viral entry mechanism(s) into PBMC subsets need to be explored.

• Addressing these important questions could help in designing effective anti-virals and recommendations on inclusion of PBMC/whole blood for diagnostic testing.

Funding

C.S. Coffin is supported by the Canadian Institutes of Health Research (CIHR).

Disclosure statement

C.S. Coffin has served as a speaker, advisory board member and/or has received research support from Bristol Myers Squibb, Glaxo Smith Kline, Gilead Sciences, Janssen and Merck.

Notes on contributors

Shivali S. Joshi is a postdoctoral fellow under the supervision of Dr. Coffin at the University of Calgary. The author’s research interests include viral hepatitis in pregnancy, co-infections and vaccine immunology.

Carla S. Coffin is an associate professor of Medicine at the Cumming School of Medicine, University of Calgary and Medical Director of the Calgary Liver unit. The author’s research interests are in hepatitis B persistence with a focus on unique patient populations (pregnancy, co-infections, transplantation, Non-Alcoholic Fatty Liver Disease).