REVIEWS

Immunobiology and pathogenesis

of hepatitis B virus infection
Matteo Iannacone 1,2,3 ✉ and Luca G. Guidotti 1,2 ✉

Abstract | Hepatitis B virus (HBV) is a non-​cytopathic, hepatotropic virus with the potential to cause
a persistent infection, ultimately leading to cirrhosis and hepatocellular carcinoma. Over the past
four decades, the basic principles of HBV gene expression and replication as well as the viral and
host determinants governing infection outcome have been largely uncovered. Whereas HBV
appears to induce little or no innate immune activation, the adaptive immune response mediates
both viral clearance as well as liver disease. Here, we review our current knowledge on the
immunobiology and pathogenesis of HBV infection, focusing in particular on the role of CD8+
T cells and on several recent breakthroughs that challenge current dogmas. For example, we now
trust that HBV integration into the host genome often serves as a relevant source of hepatitis B
surface antigen (HBsAg) expression during chronic infection, possibly triggering dysfunctional
T cell responses and favouring detrimental immunopathology. Further, the unique haemodynamics
and anatomy of the liver — and the changes they frequently endure during disease progression
to liver fibrosis and cirrhosis — profoundly influence T cell priming, differentiation and function.
We also discuss why therapeutic approaches that limit the intrahepatic inflammatory processes
triggered by HBV-​specific T cells might be surprisingly beneficial for patients with chronic infection.

Cirrhosis
The final stage of fibrosis
in which fibrous septa
surrounding nodules of
regenerating hepatocytes
induce profound architectural
distortion of the liver and
functional insufficiency.
1
Division of Immunology,
Transplantation and
Infectious Diseases, IRCCS
San Raffaele Scientific
Institute, Milan, Italy.
2
Vita-​Salute San Raffaele
University, Milan, Italy.
3
Experimental Imaging
Center, IRCCS San Raffaele
Scientific Institute,
Milan, Italy.
✉e-​mail:
iannacone.matteo@hsr.it;
guidotti.luca@hsr.it
https://doi.org/10.1038/
s41577-021-00549-4
Hepatitis B virus (HBV), the prototypical member of the
Hepadnaviridae family, is a non-​cytopathic DNA virus
that is transmitted by contacts with infected blood and
body fluids and triggers immune-​mediated liver diseases
of varying severity and duration1. The infectious virion
is an enveloped nucleocapsid that selectively enters the
hepatocyte and delivers an incomplete circular DNA
genome, thereby initiating a multifaceted process of viral
replication (Box 1; Figs 1,2).
According to WHO (World Health Organization)
data, almost one-​third of the world’s population has been
infected by HBV at some point in their lives2–4. The vast
majority of these people encountered HBV in adulthood,
developed a self-​limited infection (that is, acute hepa-
titis B (AHB)) and successfully controlled the virus2–4.
Although fewer than 5% of individuals who encountered
HBV as immunocompetent adults develop a persistent
infection (that is, chronic hepatitis B (CHB)), most of
the infections acquired in infancy or early childhood
become chronic2–4. CHB affects more than 250 million
individuals worldwide and almost 1 million die annually
from complications of persistent infection, liver cirrhosis
and hepatocellular carcinoma (HCC)2–4. CHB is highly
endemic in Africa and Asia, as well as parts of Central
and Eastern Europe2–4.
Preventive measures, including a prophylactic vac-
cine already adopted in all developed nations, remain of
paramount importance for HBV control2–4. Unfortunately,
the vaccine is ineffective in patients who are already
chronically infected, and current therapy for these
individuals is limited to either immunomodulatory
agents — such as conventional or pegylated type I inter-
ferons — or a few direct-​acting antivirals (DAAs) such
as entecavir, tenofovir and tenofovir alafenamide2–4.
These DAAs, referred to as third-​generation nucleos(t)
ide analogues (NUCs), represent the backbone of treat-
ment in most countries and — as orally administered
and well tolerated drugs that potently curtail the activity
of the HBV polymerase with extremely low resistance
rates — safely and efficiently maintain the viraemia of
patients with CHB below the detection limit of commer-
cial assays2–4. Similar to anti-​HIV drugs, however, NUCs
rarely achieve complete viral elimination (that is, HBV
eradication from the liver), and most patients therefore
require lifelong treatment2–4.
Based on the above, and coupled with the fact that
NUC treatment does not completely eliminate the risk of
developing HCC5, it is not surprising that considerable
resources are directed at developing ‘curative’ treatments
for CHB. The scientific community has long recognized
that a ‘sterilizing’ cure — the permanent elimination of
all HBV forms, including the HBV transcriptional tem-
plate (an episomal ‘minichromosome’ termed covalently
closed circular DNA (cccDNA)) (Box 1), from all infected

NATuRE REVIEWS | Immunology

0123456789();:
Reviews

Box 1 | HBV tropism and life cycle

It has been long known that hepatitis B virus (HBV) infects a
only humans, chimpanzees and, to a lesser extent, tree
shrews (Tupaia belangeri) and that the parenchymal cell of LSEC
the liver (the hepatocyte) is the sole site of productive HBV
infection137,138. Through the engagement of specific loops in HSPG
the envelope proteins, circulating virions initially attach to
heparan sulfate proteoglycans (HSPGs)139,140 that extend from
the space of Disse — which separates hepatocytes from liver Space of Disse
sinusoidal endothelial cells (LSECs) — into the lumen of the
sinusoids141 (see the figure). The following step of viral entry
involves the interaction of a specific domain of the HBV L
envelope protein with a high-​affinity receptor on the surface
of the hepatocytes. The nature of this receptor remained ssDNA rcDNA MVB
NTCP
elusive until the laboratory of Wenhui Li in 2012 elegantly or HBV
Capsid virion
identified the sodium taurocholate co-​transporting assembly
polypeptide (NTCP) as the surface molecule allowing HBV
Core Pol dsIDNA
entry into the hepatocyte142. NTCP is a hepatocyte-specific
transporter of bile acids that is predominantly localized in Capsid
disassembly pgRNA Env
the basolateral membrane that faces the sinusoidal lumen143. Filament Sphere
ER SVP SVP
Following viral entry (a process involving still unclear
phases that include membrane fusion), HBV nucleocapsids
containing a relaxed circular DNA (rcDNA) are transported to mRNA
the nucleus, where host enzymes then repair the viral genome cccDNA RNA Precore
into an episomal transcriptional template, the covalently HBeAg
closed circular DNA (cccDNA) . This circular template
10
Nucleus
encodes six capped, polyadenylated and overlapping RNAs Hepatocyte
(Fig. 2) that leave the nucleus unspliced and produce the viral
structural and non-structural proteins10. Part a of the figure b
shows how one of these transcripts — termed pre-​genomic
RNA (pgRNA) — is selectively encapsidated together with the
viral polymerase (Pol) in the hepatocellular cytoplasm10. HBV
replication takes place within these nucleocapsids by pgRNA
reverse transcription, which ultimately produces either
rcDNA or double-​stranded linear DNA (dslDNA) replicative
intermediates, depending on whether a specific event of
RNA primer translocation occurs or not144. Nucleocapsids
containing canonical (rcDNA) or non-​canonical (dslDNA)
genomes traffic back to the nucleus to amplify the pool
of cccDNA molecules or proceed to multivesicular bodies NTCP
(MVBs), where they engage the viral envelope proteins, Capsid
disassembly
bud into the lumen and exit the hepatocyte as infectious pgRNA
virions circulating in the bloodstream10,145. Env Filament Sphere
Part b of the figure shows how HBV nucleocapsids SVP SVP
containing non-​canonical dslDNA replicative intermediates,
once delivered intranuclearly, can serve as substrates for
host genome integration occurring at double-​stranded DNA mRNA
breaks. Integrated dslDNA cannot support viral replication ER
Integration
but allows the expression of specific gene products,
particularly the envelope proteins (Env) that differently
decorate filamentous and spherical subviral particles (SVPs).
SVP morphogenesis occurs at the level of the endoplasmic reticulum (ER), likely starting from the self-​assembly of the S protein into filaments; most
filaments are thought to be converted into spherical SVPs before secretion. HBeAg, HBV precore protein; ssDNA, single-​stranded DNA.

hepatocytes — will be difficult to attain in the foresee-
able future6,7, particularly given that complete cccDNA
elimination apparently does not even occur following
self-​limited AHB8. In its place, the concept of ‘functional’
cure — the loss of detectable serum hepatitis B surface
antigen (HBsAg) with or without seroconversion to
anti-​HBs antibody — has taken hold. It is notable that
each HBV envelope protein (L, M and S) possesses the
HBsAg determinant (Fig. 1) and that HBsAg circulates
in the blood of patients infected with HBV mostly in
the form of non-​infectious subviral spheres and fila-
ments, which are present in a 102-​fold to 105-​fold excess
over virions and reach exceptional quantities of up to
>300 μg ml–1 (ref.9). The reason why the virus displays
this extraordinary level of biosynthetic effort is not well
understood, but a likely explanation is that it allows
for the absorption of circulating anti-​HBsAg antibod-
ies that would otherwise limit the spread of infectious
virions within the liver. Similar to what is observed
in patients with resolved AHB, the serum HBsAg
loss associated with the concept of a ‘functional’ cure
should serve as a surrogate marker for immune control
and long-​term suppression of HBV replication. This
may represent a more achievable target than complete

www.nature.com/nri

0123456789();:

HBV elimination and could allow for cessation of
NUC treatment6,7.
All immunomodulators and DAAs that have received
approval thus far stemmed from the work of many dif-
ferent laboratories over the last four decades. Indeed,
through the study of patients, animal models and cell
cultures, the basic principles of HBV gene expression
and replication have been unveiled, infectious viral
genomes and derivative gene products have been
cloned, sequenced and characterized10 and the main
viral or host determinants governing the outcome of
infection (AHB versus CHB) have been uncovered11.
Polymerase
Core HBcAg
Large envelope
Middle envelope HBsAg
Capsid Capsid section Small envelope
42 nm
Hepatitis B virion
(Dane particle)
~90% ~10% ~1%
rcDNA dsIDNA RNA
Fig. 1 | HBV particle. Circulating hepatitis B virions (Dane particles) are ~42 nm in
diameter and are equipped with an envelope accommodating three in-​frame viral gene
products termed large (L), middle (M) and small (S) envelope proteins, each containing
the hepatitis B surface antigen (HBsAg) determinant. The hepatitis B virus (HBV) envelope
encloses an inner nucleocapsid particle (~28 nm in diameter) that is composed, in the vast
majority of virions, of 120 core protein (also known as HBcAg) homodimers assembled
into an icosahedral structure with a T = 4 symmetry. In turn, the nucleocapsid particles
contain a copy of a relaxed circular partially double-​stranded DNA genome of ~3.2 kb,
and a copy of the viral polymerase. In place of relaxed circular DNA (rcDNA), ~10% of the
nucleocapsid particles are thought to contain double-​stranded linear DNA (dslDNA) and
~1% of the particles appear to contain mostly unspliced forms of HBV RNA.
NATuRE REVIEWS | Immunology
0123456789();:
Reviews
More specifically, experimental evidence demonstrated
that during self-​limited AHB the highly non-​cytopathic
HBV induces little or no innate immune responses and
is controlled (without being completely eliminated) by
HBV-​specific CD4+ and CD8+ T cell responses as well as
neutralizing antibodies11,12. Additional work indicated
that viral persistence and chronic liver injury mostly
reflect the dysregulation of these adaptive immune
responses11,12, which are also affected by the unique
anatomy and haemodynamics of the liver (Box 2).
Here, we discuss the main findings of the work car-
ried out over the last 40 years and focus on numerous
recent breakthroughs that reveal that some virologic and
immunologic assumptions from the past are likely too
simplistic. For example, it was recently shown that defec-
tive and replication-​incompetent HBV linear fragments
integrate into the host genome early upon infection13–15
and may serve as the relevant source of HBsAg in the
context of chronic infection16 (Box 3). The extent to
which this event triggers dysfunctional T cell responses,
favours detrimental immunopathology and/or influ-
ences therapeutic strategies, will be discussed. Along
these lines, a present dogma in HBV immunobiology
asserts that high levels of circulating HBsAg negatively
impact both HBV-​specific B and T cells, and, therefore,
strategies that reduce serum HBsAg and lead to serocon-
version could result in the ‘awakening’ of these otherwise
dysfunctional responses. New experimental evidence
that challenges the validity of this dogma will be also dis-
cussed. Further, we review recent work that reveals how
anatomic changes associated with advanced liver fibro-
sis and cirrhosis may affect the priming, differentiation
and effector functions of naive or antigen-​experienced
T cells. Finally, we discuss the provocative hypothesis
that patients with CHB might benefit from therapeutic
approaches that limit the pathogenic potential of
HBV-​specific T cells. One example of such an approach
is anti-​platelet therapy (for example, with low-​dose aspi-
rin), which seems to reduce the risk of HCC and related
mortality17 by mechanisms that may include the capacity
of platelets to support the homing of pathogenic T cells
into the liver18,19.
The early phase of infection
It has long been known that, upon entering the circula-
tory system, HBV and related hepadnaviruses retain an
extraordinary capacity to reach and infect parenchymal
liver cells (Box 1). Indeed, intravenous inoculation with
just one duck HBV or HBV virion in ducks or chimpan-
zees can result in a productive infection20,21. Anatomic
and haemodynamic characteristics of the liver micro-
circulation, such as the protrusion of heparin sulfate
proteoglycans (HSPGs) in the lumen of sinusoids and
the slow blood flow that typifies this vascular bed22,23,
are thought to increase the likelihood of HBV eventually
engaging its receptor on the hepatocellular membrane
and initiating the processes of viral entry and replication
(Box 1). HBV starts replicating and spreading throughout
the liver quite rapidly, with an estimated doubling time
of 2–4 days21,24. Notably, the main features of the early
phase of infection have been best studied in biopsies col-
lected on a weekly basis from young adult chimpanzees
Reviews
PreS2
1 S
eS
Pr
b
2.1 k
e
as
er
kb
ym
2.4
Pol
Core
AA AA
AA
AA AA
AA AAA A
kb
A AA
0.7
3.5
A
kb kb
3.9
Pr
eC
X this non-​cytopathic virus to transmit to others might be
Fig. 2 | genomic structure of HBV. The inner helixes represent the partially double-​
stranded, relaxed circular hepatitis B virus (HBV) DNA genome. The outer thin lines
represent the six capped, polyadenylated and overlapping RNAs. Of note, the depicted
3.5-​kb RNA represents two transcripts 3.5 kb in length that differ from one another by
only a few nucleotides. The slightly longer transcript encodes for the precore protein
(PreC); the slightly shorter one encodes for both the core protein and the polymerase
protein. The 3.9-​kb RNA is thought to represent a long form of a HBV xRNA; note that
the canonical HBV xRNA is 0.7 kb in length. The 2.4-​kb RNA encodes for the PreS1 (or L)
envelope protein. The 2.1-​kb RNA encodes for both the PreS2 and S envelope proteins
(also called M and S). The outermost colour lines indicate the translated HBV proteins.
that were inoculated intravenously with transgenic
mouse serum containing varying amounts of the same
infectious HBV DNA (HBV genotype D subtype ayw)25.
These studies revealed that, depending on the inoculum
dose, HBV easily manages to infect 100% of the hepat-
ocytes (~1011 cells in the chimp or human liver)23 within
a period of 7–10 weeks and this occurs without any evi-
dence of hepatocellular damage21,26–28. Interestingly, the
outcome of infection was found not to strictly depend on
the inoculum dose — meaning that a chronic infection
with associated immunopathology can also result from
the inoculation of a very low HBV dose — but, rather,
on the relationship between the kinetics of viral spread
and the priming of adaptive immunity, particularly that
of HBV-​specific CD4+ T cells21. These studies in chim-
panzees also uncovered that, during the initial spreading
phase, HBV is highly efficient at avoiding recognition
by the innate immune system26–28, possibly owing to its
unique replication strategy that involves, for instance,
the nuclear sequestration of the transcriptional template
0123456789();:
(that is, the cccDNA molecule), the use of capped and
polyadenylated transcripts resembling host-​derived
mRNAs and the confinement of replicating RNA/DNA
genomes within cytoplasmic nucleocapsid particles10.
Interestingly, more recent work on human liver biop-
sies indicated that HBV seems not to elicit detectable
innate immune responses in the CHB setting either29.
Along with the observation that different aspects of the
HBV life cycle are highly sensitive to antiviral molecules
produced by experimentally activated innate immune
cells such as intrahepatic antigen-​presenting cells, natu-
ral killer cells and natural killer T cells30–38, these results
suggest that — instead of actively counteracting the
effect of such molecules as many other viruses do39 —
HBV has evolved a hiding strategy to avoid recognition
by the innate immune system, which allows it to spread
and replicate proficiently within the liver.
Regarding the studies in chimpanzees, it is note-
worthy that at the same time points at which the
hepatic HBV titres and the percentage of infected
hepatocytes reach their peak, viraemia can be as high as
1011 genomes ml–1 and, in keeping with the lack of innate
immune cell-​derived pro-​inflammatory mediators, this
occurs in animals showing no clinical, biochemical or
histological evidence of liver disease21,26–28. Note that sim-
ilar levels of viraemia have been documented in patients
during the incubation phase of infection, at least 4 weeks
before the onset of symptoms40. It therefore appears that
at times of maximum viral production, the capacity of
very high (a nanolitre of blood from a clinically ‘healthy’
individual may contain up to 105 genomes).
Priming of adaptive immune responses in immuno­
competent adults infected with HBV. Human and animal
studies over the last 30 years have indicated that the out-
come of most adult infections is strongly determined
by the kinetics, breadth, vigour, trafficking and effec-
tor functions of HBV-​specific adaptive immune cells1.
Through their antiviral functions, such as the killing of
infected hepatocytes and local production of antiviral
cytokines, HBV-​specific CD8+ T cells are widely con-
sidered the ultimate effectors of viral clearance during
AHB, whereas HBV-​specific CD4+ T cells are viewed as
essential facilitators of the induction and maintenance of
both CD8+ T cell and antibody responses1. Accordingly,
the depletion of either CD8+ or CD4+ T cells during the
initial phase of infection in acutely infected chimpan-
zees prevented both viral clearance and the onset of
liver disease21,41. It is of note that, in chimpanzees, anti-
bodies with neutralizing potential appear only after the
virus has been eliminated from most hepatocytes21,26–28,
suggesting that their main function is to prevent the
re-​emergence of HBV following clinical recovery. This
stands in contrast to the HBcAg-​specific IgM (anti-​HBc)
observed early in infection, which can coexist with
high levels of HBV replication42. The function of these
antibodies in HBV pathogenesis is currently unclear43.
We still know very little about the timing and the
location of T cell priming in both AHB and CHB. In
the setting of adult infection, and in keeping with data
derived from experiments in chimpanzees, HBV-​specific
www.nature.com/nri
 Reviews

Cross-​priming
A functional outcome of
cross-​presentation (the
presentation of extracellular
antigens on MHC class I
molecules), whereby
antigen-​specific naive CD8+
T cells are activated by
antigen-​presenting cells
to become effector cells.
CD8+ T cells have been detected during the initial
asymptomatic phase of AHB, typically no sooner than
9–11 weeks after infection21,26,40. The presence of these
cells is often kinetically associated with non-​cytolytic,
cytokine-​dependent antiviral events that were first
described in HBV transgenic mice 44,45. Why the
HBV-​specific effector CD8+ T cells cannot be detected
at much earlier time points (for instance, 2 or 3 weeks
after viral exposure when HBV could have infected only
a minority of the hepatocytes) remains to be determined.
The immunological dogma would dictate that naive
CD8+ T cells initially encounter their cognate antigen in
secondary lymphoid organs (SLOs), where cross-​priming
by professional antigen-​presenting cells of circulating
virions and subviral particles promotes the differentia-
tion into effector cells endowed with liver homing poten-
tial. Yet there is no direct experimental evidence that
SLOs are the priming site for CD8+ T cells in humans
or chimpanzees infected with HBV, and it is assumed
that intrahepatic priming of naive CD8+ T cells (espe-
cially in the presence of high antigen load) promotes the
functional impairment of these cells46. Based on these
considerations, it is generally believed that the priming
of HBV-​specific naive CD8+ T cells in the liver does not
promote viral clearance but, rather, contributes to the
establishment of persistent infection.
Recent experiments in mouse models of HBV infec-
tion have indeed demonstrated that hepatocellular prim-
ing of naive HBV-​specific CD8+ T cells leads to their
local activation and proliferation, but they do not dif-
ferentiate into effector cells47,48 (Fig. 3). These studies also
revealed that the CD8+ T cell dysfunction induced by
hepatocellular priming can be counteracted in vivo
by treatment with IL-2 (ref.48) (Fig. 3). IL-2 was previously
shown to rescue CD8+ T cells from death induced by
in vitro priming by hepatocytes49 and different formula-
tions of IL-2 have been clinically used as therapeutics to
increase T cell responses against viruses or tumours50–52.
The cellular and molecular mechanisms whereby
IL-2 reinvigorates intrahepatically primed T cells and
whether IL-2 plays a role during natural HBV infection
remain elusive. However, based on the aforementioned
results48, one could envisage that, during AHB, high local
IL-2 concentrations might render the liver fully capable
of supporting efficient HBV-​specific CD8+ T cell prim-
ing. A possible source of IL-2 in this setting is effector

Box 2 | The organ: how the liver influences HBV immunobiology and pathogenesis
The liver is composed of functional units (that is, lobules) where plates of hepatocytes one cell thick are in close contact with the
fenestrated and basement membraneless liver sinusoidal endothelial cells (LSECs) that form the organ microcirculation146,147
(see the figure). Hepatocytes and LSECs are separated by the space of Disse where about half of the lymph in the body is
formed148. The liver microenvironment also includes epithelial cells of the bile ducts termed cholangiocytes, intrasinusoidal
macrophages termed Kupffer cells, perisinusoidal myofibroblasts termed hepatic stellate cells and selective populations
of resident cells of lymphoid or myeloid origin that constantly patrol liver sinusoids or reside within periportal areas147,149.
Under both physiologic and pathologic conditions, the liver is biased towards inducing a state of immune unresponsiveness
or hypo-​responsiveness. This involves a complex array of diverse and coordinated cellular and molecular events that,
ultimately, hinder the effector functions of lymphocytes147,149,150. For instance, the unique anatomy and haemodynamics of
the liver sinusoids — through which about one-​third of all blood cells transit slowly every minute23,90 — allow circulating T cells
to sense tolerogenic MHC–Ag complexes and other surface ligands displayed by non-​professional antigen-​presenting cells
such as the hepatocytes without the need for extravasating89,147,149. Such engagement promotes the induction and maintenance
of immune tolerance, explaining, for example, the acceptance of liver allografts across complete MHC mismatch barriers151
or the propensity of hepatitis B virus (HBV) to establish lifelong persistent infections1.

Bile duct Cholangiocyte Hepatocyte Bile canaliculi

Hepatic artery

LSEC Fenestrations

Sinusoid Kupffer cell Central vein

Lymphatic
vessel Space of Disse

Hepatic
stellate cell

Portal vein

NATuRE REVIEWS | Immunology

0123456789();:
Reviews
Box 3 | HBV integration into the host genome
Hepatitis B virus (HBV) and other closely related hepadnaviruses (such as duck HBV
and woodchuck hepatitis virus) tend to extensively integrate into the host genome10
(Box 1). The mechanisms governing this phenomenon are poorly defined but it is clear
that non-​canonical double-​stranded linear DNA (dslDNA) replicative intermediates,
once delivered intranuclearly, can serve as substrates for host genome integration
occurring at double-​stranded DNA breaks through non‑homologous end joining or
microhomology-​mediated end joining10. HBV integration can start as early as <3 days
after infection in cell cultures13 and rapid integration events have long been known
to happen in vivo as well14,15,152. The process of HBV integration has the potential to
endlessly continue during persistent infection, but cannot support viral replication.
This is because even intact forms of integrated 3.2-​kb dslDNA cannot generate the
3.5-​kb pre-​genomic RNA (pgRNA) from which the replication of HBV actually starts.
Indeed, pgRNA synthesis requires a 3.2-​kb circular genome on which the viral polymerase
recognizes the unique polyadenylation signal not at the first but at the second passage10
(Fig. 2), and the redundant sequences that the polymerase uses to produce pgRNA are
simply missing in dslDNA forms. Nonetheless, forms of integrated dslDNA can support
the expression of specific gene products. The virologic and tumorigenic implications of
this phenomenon — which include the production of amino-​terminal truncated versions
of the HBV X protein (a transactivator that regulates viral transcription from covalently
closed circular DNA (cccDNA)) or the generation of HBV–cellular fusion transcripts that
yield chimeric proteins with oncogenic potential — have been expertly reviewed
elsewhere119,153,154. Notably, the open reading frames (ORFs) encoding the viral polymerase,
the precore protein (a non-​structural gene product whose mature form is secreted in the
bloodstream as HBV precore protein (HBeAg)) and the core protein (the structural and
non-​secretable component of HBV nucleocapsids also referred to as HBcAg) detach from
their promoters in dslDNA genomes153. As such, HBV integrants may produce HBeAg or
HBcAg only if properly inserted near active cellular enhancers/promoters and, to our
knowledge, definitive information about this possibility or its related frequency during
HBV infection is lacking. A different scenario pertains to the ORFs encoding the L, M and
S envelope proteins, which can remain properly positioned under their native promoters
upon dslDNA integration153 (Box 1).
It has long been suggested that integration leads to hepatic hepatitis B surface
antigen (HBsAg) expression in patients with chronic hepatitis B (CHB)155 and that
such expression can be significantly dysregulated in vitro or in vivo — for instance,
with abnormal production and accumulation of the L envelope protein and/or specific
L or M mutants in the endoplasmic reticulum and the formation of ground glass
hepatocytes156–158. Ground glass hepatocytes are indeed historical hallmarks of CHB159
and may represent preneoplastic foci of HBV-​related hepatocellular carcinoma
(HCC)158. Taking advantage of patients and chimpanzees with chronic infection
undergoing RNA interference (RNAi)-based anti-​HBV treatment, Wooddell et al.
recently made a challenging observation: HBV integrants, rather than cccDNA, seem
to support most of the hepatic HBsAg biosynthesis, particularly in individuals with
long-term infection who become serum HBeAg-​negative16. This observation, coupled
with the evidence that HBV integration can start rapidly after infection and that large
hepatocyte clones holding integrated HBV DNA can be detected during all stages of
CHB160, may force us not only to revise the way in which spontaneous or therapy-​induced
changes in serum HBsAg (or the lack thereof) have been interpreted thus far but also to
reflect on the immunological and pathological consequences of having hepatocellular
HBsAg continuously produced and presented to T cells or B cells independently of
viral replication.
CD4+ T cells; in support of this hypothesis are the obser-
vations that CD4+ T cell depletion in chimpanzees prior
to HBV infection prevents effective CD8+ T cell prim-
ing and leads to persistent infection with negligible
immunopathology21 and that detection of CD8+ T cells
in the liver of chimpanzees infected with HBV coincides
with the detection of CD4+ T cells in the liver26. Without
the creation of a manipulable animal model of infection
Checkpoint inhibitor (such as an immunocompetent mouse engineered to
therapy support HBV entry, cccDNA formation and productive
A form of cancer
immunotherapy targeting
viral replication in the liver), it will be difficult to provide
immune checkpoints (for definitive insights into when or where HBV priming
example, PD1, CTLA4). occurs following the initial exposure.
0123456789();:
Priming and other immune correlates of neonatal and
perinatal infections. As mentioned above, the vast
majority of neonatal and perinatal HBV infections
become persistent1. This outcome is thought to be due
to immune inhibitory mechanisms that — depending
on the relative functional avidity of HBV-​specific T cell
clones for tolerogenic epitopes — either delete these
cells in the thymus or render them dysfunctional in the
periphery53. However, direct evidence of thymic deletion
of T cell clones following viral exposure is lacking.
Moreover, circulating HBV-​specific T cells are detecta-
ble in adolescents and young adults infected neonatally
and perinatally (albeit at levels that are both quantita-
tively and qualitatively deficient when compared with
those characterizing self-​limited infections acquired in
adulthood)54, and HBV-​specific T cell clones, although
tolerized, are still present in fully grown transgenic mice
that were exposed to HBV antigens at birth55,56. These
observations seem to argue against a predominant role
of thymic deletion of HBV-​specific T cells in the estab-
lishment of viral persistence. Considering that naive
CD8+ T cells emerging from the thymus should easily
engage hepatocellular antigens in the poorly inflamed
liver, and that HBV robustly replicates during the first
20–30 years of life in patients with CHB exposed neo­
natally/perinatally54, it is plausible that tolerogenic
events that result from the initial antigen encounter in
the liver may impact the outcome of infection.
As mentioned above, recent experimental evidence in
mouse models of HBV infection showed that hepatocel-
lular priming of naive HBV-​specific CD8+ T cells leads
to local activation and proliferation of these cells but a
lack of differentiation into effector cells47,48. These dys-
functional cells form clusters that coalesce around portal
tracts48 (Fig. 3), similar to what is observed in patients
with CHB57. This periportal accumulation happens even
though the initial antigen recognition by naive CD8+
T cells occurs homogeneously throughout the liver
lobule48. Hepatocellular HBV antigen recognition by
naive CD8+ T cells initiates a differentiation programme
characterized by a progressive accumulation of tran-
scriptional and chromatin changes that, ultimately, result
in a dysregulated T cell phenotype48. Genes involved in
tissue development, remodelling, cell differentiation
and wound healing are upregulated48, suggesting that
this programme might favour tissue protection over
antiviral activity. The acquisition of this dysfunctional
phenotype appears to not be affected by the relative anti-
gen load in the hepatocyte (even very small amounts of
HBV antigens can trigger severe T cell dysfunction)48.
Interestingly, the transcriptional signature of CD8+
T cells primed by hepatocytes does not overlap signif-
icantly with that of exhausted T cells described in other
viral infections and in cancer48. Consistent with this,
hepatocellularly primed CD8+ T cells do not respond
to anti-​PDL1 checkpoint inhibitor therapy48. Whereas
priming by hepatocytes in a non-​inflamed liver (a likely
scenario following neonatal HBV infection) triggers
this unique dysfunctional gene programme, antigen
persistence may gradually activate an exhaustion-​like
gene signature48,58. These considerations should guide
the design and interpretation of ongoing clinical trials
www.nature.com/nri
 Reviews

aimed at evaluating the therapeutic potential of immune
checkpoint inhibitors in patients with CHB.
Unfortunately, very few immunological studies have
been performed in experimentally infected chimpan-
zees that eventually developed CHB59, and none of them
investigated whether T cell priming can occur in the
liver. However, one study indicated that HBV-​specific
CD4+ T cell responses are not detectable in these ani-
mals before the peak of viraemia or before the virus
has infected most hepatocytes21. As the hepatic appear-
ance of HBV-​specific CD8+ T cells is also significantly
delayed in relation to those time points, it seems — as
mentioned earlier — that the late priming of CD4+ T cell
responses fails to support the development of quanti-
tively and qualitatively effective CD8+ T cell responses,
thus favouring viral persistence21.
Whatever the relative contribution of central or
peripheral tolerance to viral persistence, it is noteworthy
that viral factors such as the HBV precore protein
(HBeAg) — a viral protein that has potential to cross the
placenta and is not required for viral assembly, replica-
tion or infection10 — have long been suggested to pro-
mote these tolerogenic events. Indeed, mouse models of
HBV infection have demonstrated that HBeAg can act
as a tolerogenic protein for HBcAg and HBeAg-​specific
T cells60,61 and, more recently, that maternally derived
HBeAg appears to condition the hepatic macrophages
of the offspring to repress HBV-​specific CD8+ T cell
responses by yet undetermined mechanisms 62,63.
In keeping with a role for HBeAg in blunting T cell
responses, it has long been known that neonatal infec-
tions with HBeAg-​negative variants often result in
viral clearance and that adult infections with these
variants are usually more clinically severe than those
with HBeAg-​c ompetent genomes 64. In addition to
HBeAg, HBsAg — which is abundantly produced by
either cccDNA or integrated HBV forms — could serve
as a high-​dose tolerogen during CHB. The fact that
envelope-​specific T cell responses are usually very dif-
ficult to detect in patients with CHB — as compared,
for instance, with those specific for the core or the poly-
merase proteins65,66 — fits with this hypothesis. As stated
above, it is generally acknowledged that the relatively
high levels of circulating HBsAg that characterize most
CHB infections negatively influence both virus-​specific
B and T cell responses and that the clearance of serum
HBsAg might restore dysfunctional responses. However,
recent data obtained in transgenic mice and from
patients challenge this assumption67,68. In mice, for
instance, serum HBsAg loss does not alter naive CD8+

HBV-specific
naive CD8+ T cell

Kupffer cell priming Hepatocellular priming Hepatocellular priming + IL-2

IL-2

CD8 Kupffer cell

TCR
MHC class I

• Bona fide effector cells • Genes of tissue remodelling • Rescuing of effector genes
• Parenchymal clusters • Periportal clusters • Parenchymal clusters
• Effector functions • Dysfunctional phenotype • Effector functions

Fig. 3 | Spatiotemporal dynamics and genomic landscape of CD8+ T cells undergoing intrahepatic priming. Priming
of naive CD8+ T cells by Kupffer cells — which are not natural targets of hepatitis B virus (HBV) — in mouse models leads to
the generation of bona fide CD8+ effector T cells that accumulate in clusters scattered throughout the liver parenchyma
(left). Priming of naive CD8+ T cells by hepatocytes induces their local activation and proliferation but fails to induce the
differentiation into effector cells. The resulting dysfunctional cells show little transcriptional overlap with exhausted
T cells, upregulate genes associated with wound healing/tissue remodelling and accumulate in clusters around portal
tracts (middle). The CD8+ T cell dysfunction induced by hepatocellular priming can be overcome by in vivo IL-2 administration,
which appears to rescue the expression of effector genes48 (right).

NATuRE REVIEWS | Immunology

0123456789();:
Reviews
SSB/La-​dependent
mechanism
T cell-​induced cytokines
such as IFNγ and TNF have
been shown to induce
the post-​transcriptional
downregulation of hepatitis B
virus (HBV) RNAs in vivo. This
process appears to rely on the
degradation of the full-​length
SSB/La protein, which normally
functions as a HBV RNA
stabilizer in the nucleus of the
hepatocyte.
T cell fate upon intrahepatic priming and does not alter
their response to an IL-2-​based immunotherapy67. This
suggests that enhanced virus-​specific T cell responses in
patients with CHB might actually cause (rather than be a
consequence of) HBsAg clearance by killing or ‘curing’
HBsAg-​producing hepatocytes. Monoclonal antibody
therapy targeting HBsAg is currently being evaluated in
patients with CHB69–72. However, these results suggest
that antibody therapy alone, although it might reduce
the cccDNA pool possibly maintained by the infection
of new hepatocytes7, is unlikely to improve HBV-​specific
CD8+ T cell responses.
In addition to specific viral proteins, viral mutations
may also contribute to T cell impairment. Although
definitive evidence that HBV can mutate, and thereby
escape pathogenic T cell responses, is yet to be pro-
duced, mutations in epitopes that ultimately antagonize
T cell recognition or induce anergy in T cells have been
described in patients with CHB73–75. Moreover, several
other non-​viral physiologic or disease-​driven tolero-
genic mechanisms that affect not only the priming but
also the expansion and effector function of HBV-​specific
T cells have been proposed. These include the produc-
tion of high levels of soluble mediators (such as argin-
ase, indoleamine 2,3-​d ioxygenase and suppressive
cytokines), the upregulation of inhibitory checkpoint
receptor/ligand pairs and the expansion of regulatory/
suppressive cells and natural killer cells53,76,77. Indeed,
multiple primary and secondary immune mechanisms
of HBV hypo-​responsiveness are likely to simultaneously
operate in the liver and/or SLOs of individuals infected
neonatally and perinatally (and, possibly, the liver and/or
SLOs of the few individuals that acquire CHB in adult-
hood). These considerations call into question whether
simple immunotherapeutic approaches — tackling, for
instance, one tolerogenic mechanism at a time, as cur-
rently pursued in some current clinical studies — will
have curative potential for CHB.
In contrast to T cells, very little is known about B cell
priming in the context of neonatal and perinatal infec-
tions. In stark contrast to what is observed in individ-
uals with resolved AHB, most patients who eventually
develop CHB do not show detectable levels of anti-​HBs
antibodies43. The reason for this might be the sequestra-
tion of antibodies by large amounts of soluble HBsAg (we
still lack a widespread clinical assay for the detection of
HBsAg–HBsAb complexes), the lack of HBsAg-​specific
B cells or their functional impairment43. Recent work
showed that, when compared with HBV-​vaccinated con-
trols, patients with CHB with persistently high HBsAg
levels have similar frequencies of HBsAg-​specific B cells
but show impaired production of anti-​HBs antibodies78–81.
HBsAg-​specific B cells in patients with CHB were found
to be enriched in cells with an atypical memory pheno-
type — to the detriment of conventional, classic memory
B cells78,79. When compared with HBsAg-​specific B cells,
HBcAg-​specific B cells are present at higher frequen-
cies in patients with CHB, and these can mature into
anti-​HBc-​producing cells in vitro82,83. This difference
might be due to the lower quantity of HBcAg available
(compared with HBsAg) or on the ability of HBcAg
to activate B cells in a T cell-​independent fashion84.
0123456789();:
Effector phase during self-​limited AHB
Regardless of where priming occurs, effector CD8 +
T cells, which could facilitate viral clearance, need to
traffic within the liver and recognize hepatocellular
antigens in order to exert their antiviral functions85–88.
Recent studies in transgenic mice revealed that effec-
tor CD8+ T cells arrest in liver sinusoids, rather than in
in postcapillary venules. This process does not involve
classical leukocyte adhesion molecules (such as selec-
tins, chemokines and integrins) or antigen recognition89;
rather, CD8+ T cells bind to small platelet aggregates that
form when platelet-​expressed CD44 interacts with sinu-
soidal hyaluronic acid89. Upon this initial arrest, effector
CD8+ T cells crawl at 10 μm min–1 (ref.89) (~1,000-​fold
slower than the sinusoidal blood flow90) and recognize
hepatocellular antigens while still intravascular89. This
occurs via cytoplasmic protrusions that extend through
sinusoidal endothelial fenestration89 (Fig. 4). It is only after
deploying effector functions (including the production
of IFNγ or the killing of HBV-​expressing hepatocytes)
that effector CD8+ T cells extravasate89, a process that
might have evolved to limit excessive immunopathology
(note that hepatocytes express MHC class I mostly on
the membrane side that faces the sinusoidal lumen)91.
As mentioned above, hepatocellular recognition by
HBV-​specific effector CD8+ T cells eventually leads to
antiviral cytokine production and hepatocyte killing92–94.
The latter is an intrinsically inefficient process, demand-
ing physical contact between the infected hepatocyte
and the T cell. It could therefore be difficult for effector
CD8+ T cells to reach and kill each infected hepatocyte,
particularly if one considers that all 1011 hepatocytes
are usually infected during self-​limited AHB in chim-
panzees and relatively few virus-​specific effector CD8+
T cells are present in the blood and liver of these ani-
mals. Further, should the rapid clearance of HBV merely
require the killing of most hepatocytes, one would expect
these acutely infected chimpanzees to show a much
higher degree of liver disease severity than is actually
observed26. Along with these considerations, a large body
of experimental evidence in HBV replication-​competent
transgenic mice demonstrated that the inhibition
of viral replication by effector CD8+ T cells is mainly
dependent on non-​cytolytic mechanisms that involve
the local production of IFNγ and TNF by these cells26,45.
As already mentioned, the initial phase of viral clearance
in chimpanzees kinetically coincides with the intra­
hepatic appearance of IFNγ and TNF26. It is of note that
these cytokines, possibly through their ability to induce
nitric oxide in the liver95, appear to prevent the assem-
bly of replication-​competent HBV RNA-​containing
capsids in the hepatocyte96 in a proteasome-​dependent
manner97. During this process, the HBV nucleocapsids
vanish from the cytoplasm of the hepatocytes25 and viral
RNAs are destabilized in the hepatocellular nucleus by
an SSB/La-​dependent mechanism30–32. IFNγ and TNF also
appear to affect the stability of cccDNA in the nucleus
of non-​dividing hepatocytes and, again, this occurs dur-
ing the initial phase of viral clearance in acutely infected
chimpanzees26. More recent in vitro experiments with
primary and immortalized human hepatocytes that had
previously been infected with HBV have confirmed
www.nature.com/nri
 Reviews

Effector
CD8+ T cell
Platelets
Docking on
adherent platelets Antigen recognition
and killing
Crawling

Hyaluronan
LSEC

Apoptosis

Hepatocyte HBV

Fig. 4 | Immune surveillance of the liver by effector CD8+ T cells. Effector CD8+ T cells circulating through the mouse
liver initially arrest within sinusoids, not postcapillary venules. They do so independently of antigen recognition and of
numerous molecules variably implicated in leukocyte trafficking to other organs, such as selectins, Gαi-​coupled receptors,
β2 and α4 integrins, platelet endothelial cell adhesion molecule 1 (PECAM1) and vascular adhesion protein 1 (VAP1).
Rather, the predominant mechanism for arrest of circulating effector CD8+ T cells within liver sinusoids is by docking onto
platelets that have previously adhered to hyaluronan on liver sinusoidal endothelial cells (LSECs) via the surface receptor
CD44. After the initial platelet-​dependent arrest, effector CD8+ T cells actively crawl along liver sinusoids. Binding to
hepatitis B virus (HBV) antigens presented by infected hepatocytes leads to antiviral cytokine production, and hepatocyte
killing occurs in a diapedesis-​independent manner. These processes are mediated via the extension of cellular protrusions
through sinusoidal endothelial fenestrae by effector CD8+ T cells, producing contact sites with the sub-​sinusoidal
hepatocellular membrane.

these findings98. Indeed, in these systems, IFNγ and
TNF were shown to non-​cytopathically destabilize the
nuclear pool of cccDNA via the activation of nuclear
APOBEC3 deaminases98. Together, these results suggest
that under specific circumstances (such as when exposed
to the local production of antiviral cytokines by activated
effector CD8+ T cells) this form of HBV DNA, which is
generally acknowledged to have a long half-​life, might be
eliminated from the liver rather quickly and even in the
absence of hepatocellular division. In acutely infected
chimpanzees, this phase of non-​cytopathic clearance
of HBV replicative intermediates and cccDNA usually
precedes a phase of disease-​associated cccDNA loss26,
where both the processes of hepatocellular death and
compensatory hepatocellular division could contrib-
ute to further cccDNA elimination10. According to
this scenario, the liver disease that HBV-​specific CD8+
T cells eventually trigger may be viewed as a necessary
step to complete the elimination of HBV replicative
intermediates and cccDNA from most hepatocytes. As
stated above, clearance of HBV from the liver follow-
ing AHB is rarely complete, and very small numbers
of HBV-​replicating hepatocytes are thought to persist
indefinitely in the organ8,99. Along with neutralizing anti-
bodies that are likely to prevent the re-​emergence of pro-
ductive infections from these reservoirs, it is tempting to
speculate that the long-​term presence of traces of viral
antigens may actually help to maintain T cell responses.
On this account, recent work showed that a population
of tissue-​resident memory CD8+ T cells survives —
possibly due to cell-​autonomous IL-2 production — in
the liver of individuals with resolved AHB infection (and
in the liver of patients with CHB with low viral loads)100.
It is unknown how the few HBV-​replicating hepatocytes
manage to avoid T cell recognition and persist over time.
A potential explanation may be that the virus or the
process of viral integration induces the downregulation
of molecules that are necessary to process and present
viral antigens.
HBV-​ s pecific CD8 + T cells may promote liver
immunopathology, either by directly killing infected
hepatocytes or by indirectly recruiting pathogenic
inflammatory cells into the liver. Studies in mouse mod-
els of HBV pathogenesis show that CD8+ T cells first kill
HBV-​expressing hepatocytes by rapidly inducing hepato­
cellular apoptosis via both the perforin-​dependent and
Fas-​dependent pathways101,102. Effector CD8+ T cells
initially trigger the apoptotic death of a relatively
small number of hepatocytes101,102 that are then rapidly
removed by Kupffer cells, which helps to contain liver
inflammation103. As time progresses, however, apop-
totic hepatocytes that are not promptly removed by
Kupffer cells become secondarily necrotic and release
damage-​associated molecular pattern molecules that
actively recruit neutrophils into necro-​inflammatory
foci103,104. Within these initial necro-​inflammatory foci,
which are scattered throughout the liver parenchyma,
the neutrophils become activated and release a number
of inflammatory mediators as well as matrix metallo-
proteinases (MMPs)105,106. MMPs can efficiently degrade
matrix components (including collagen, laminin and
fibronectin) that are deposited de novo (possibly by acti-
vated hepatic stellate cells) during the early process of
liver repair106. In turn, these neutrophil-​derived MMPs —
in conjunction with chemokines that are locally pro-
duced by parenchymal and non-​parenchymal liver cells

NATuRE REVIEWS | Immunology

0123456789();:
Reviews
sALT values
The serum concentrations of
the liver enzyme alanine
aminotransferase. Commonly
measured clinically as a
biomarker for liver damage.
Space of Disse
(Also referred to as
perisinusoidal space). The
space that lies between the
hepatocytes and the sinusoids.
in response to T cell-​derived cytokines107 — promote
the arrival of numerous mononuclear cells (including
HBV-​non-​specific B and T cells, as well as natural killer
cells, natural killer T cells, dendritic cells and mono-
cytes)105,106. In the end, the intrahepatic recruitment of
these inflammatory cells, which ultimately exceed the
number of HBV-​specific effector CD8+ T cells within
the lesions by more than an order of magnitude, signifi-
cantly exacerbate the severity of liver disease107. Of note,
work in patients with HBV indicates that the degree of
liver damage correlates better with the extent of hepatic
infiltration of inflammatory cells than with the frequency
of intrahepatic HBV-​specific CD8+ cells108. It should be
also noted that we know very little about the pathogenic
mechanisms whereby the abundantly recruited inflam-
matory cells induce organ damage. It is conceivable that
the local production of pro-​inflammatory and cytotoxic
mediators and/or the formation of thromboembolic
reactions in the liver microcirculation underlie this
damage, and this should be addressed experimentally.
Overall, preclinical data105–107 suggest that up to 50% of
the sALT values that are associated with immune-​active
phases of HBV infection could reflect the capacity of
inflammatory cells to kill both infected and uninfected
hepatocytes.
Effector phase during CHB
Similar to AHB, the basic principle of HBV-​specific
CD8+ T cells initially eliciting pathogenic events, with
secondary involvement of inflammatory cells, also
holds true for CHB. The fact that, when detectable,
HBV-​specific CD8+ T cells are only present at very low
frequencies in the blood and liver of patients with CHB,
and that these peripheral frequencies are usually much
lower than those of patients with AHB65,66,109–112, helps to
explain why individuals with CHB tend to develop much
less severe peaks of liver disease11,12. The probability of
detecting peripheral HBV-​specific CD8+ T cells is higher
in patients with CHB with low viraemia, and, as men-
tioned earlier, varies with the specificity of the T cells, in
that core-​specific and polymerase-​specific CD8+ T cells
are usually more easily detectable than envelope-​specific
CD8+ T cells65,66,109–112. Whether these findings reflect dif-
ferences in the intrahepatic load of the viral antigens —
with high loads promoting the deletion and/or liver
sequestration of cognate CD8+ T cells — is unknown.
Additional ex vivo and in vitro analyses of peripheral
T cells from patients with CHB indicate that core-​specific,
polymerase-​specific or envelope-​specific CD8+ T cells
are also differently impaired in their capacity to
proliferate or kill target cells, although their capacity
to produce antiviral cytokines (for example, IFNγ and
TNF) appears to be uniformly compromised66,109,111,112.
Similarly, mouse models of HBV infection revealed that
sustained antigen stimulation in vivo, as occurs when
HBV is not rapidly cleared from the liver, creates a set-
ting in which HBV-​specific CD8+ T cells lose the capa­
city to secrete IFNγ and TNF but keep some cytotoxic
activity18,113,114.
According to the data discussed above, CHB could be
viewed as a disease in which HBV-​specific CD8+ T cells
that escape central tolerance and are not completely
0123456789();:
anergized in the periphery end up attacking liver cells
in the attempt to eliminate HBV from the liver. Small
numbers of HBV-​specific CD8+ T cells that are incapable
of producing antiviral cytokines may therefore sustain
long-​term immunopathological responses without ever
achieving viral clearance. The extent to which some of
these responses — in particular, those of T cells specific
for HBsAg — are affected by HBV integration into the
hepatocyte genome and the resulting antigen presenta-
tion by hepatocytes that are either replicating HBV or
not is still poorly understood. Notably, a recent study
utilized antibodies specific for HLA/HBsAg peptide
complexes to probe liver biopsies and various cell cul-
ture systems for antigen presentation115. The findings
indicate that envelope epitopes that can be recognized
by CD8+ T cells are present, albeit not uniformly dis-
tributed, in the liver of patients with CHB and that
HBsAg derived from either cccDNA or HBV integrated
into the genome can be properly processed and pre-
sented to CD8+ T cells115. Accordingly, mouse models
of immune-​mediated pathology of CHB showed that
the persistent recognition of HBsAg produced by an
integrated transgene renders envelope-​specific CD8+
T cells fully capable to sustain prolonged, albeit mod-
erate, liver cell injury18,113,116. Although suggesting that
HBV integration may indeed have a significant impact
on the immunopathology of CHB, the latter experiments
also revealed that moderately severe but persistent CD8+
T cell-​dependent liver disease underlies the continuous
episodes of hepatocellular necrosis and hepatocellular
regeneration that, ultimately, lead to the complications
of CHB: liver fibrosis, liver cirrhosis and HCC18,113,116.
It should be noted that these complications in humans
usually take place after decades of low-​level liver disease
with associated hepatocellular regeneration (that is, cel-
lular DNA synthesis) and inflammation (that is, the pro-
duction of mutagens)1,117. Based on this preclinical and
clinical evidence, it has therefore been postulated that
the spatiotemporal coexistence of hepatocellular necro-
sis, hepatocellular regeneration and liver inflammation
may trigger abnormal repair functions and random
genetic damage, ultimately leading to liver fibrosis, liver
cirrhosis and HCC1,117. As the interval between HBV
exposure and the complications of persistent infection
is typically several decades, HBV is thought to be nei-
ther directly nor acutely oncogenic. Nonetheless, viral
factors, including the deregulation of procarcinogenic
genes due to HBV integration into the genome and/or
the expression of HBV-​derived procarcinogenic poly­
peptides, have been proposed to promote hepatocarcino­
genesis during CHB, and we refer to other reviews for
the in-​depth discussion of this topic118–120.
As mentioned above, a relevant aspect of the immune-
mediated effector phase during CHB is the development
of liver fibrosis and cirrhosis121. Of note, these pathol-
ogies are characterized by an imbalance between
fibrogenesis and fibrolysis, resulting in an intrahepatic
deposition of extracellular matrix that is qualitatively
different in its organization and composition from that
of normal liver repair122. Extracellular matrix depo-
sition in the space of Disse and reduced liver porosity
(that is, the decrease in the number and size of sinusoidal
www.nature.com/nri
 Reviews

endothelial cell fenestrae) are pathological hallmarks of
liver fibrosis and cirrhosis123. These events (referred to
as sinusoidal ‘capillarization’ and ‘defenestration’, respec-
tively) not only alter the normal exchange of soluble fac-
tors between blood and hepatocytes123 but also appear to
negatively impact T cell function in the liver89. Indeed,
recent experiments in mouse models of HBV infection
demonstrated that both liver capillarization and/or
defenestration inhibit hepatocellular antigen recogni-
tion by effector CD8+ T cells89. This suggests that the
processes of liver fibrosis and cirrhosis might reduce
CD8+ T cell-​mediated immune surveillance towards
infected or transformed hepatocytes. In addition to lim-
iting the potential of virus-​specific or tumour-​specific
CD8+ T cells for eliminating hepatocytes that are either
HBV-​infected or transformed, these anatomical and
functional perturbations of the liver microenvironment
need to be considered when selecting patients with
CHB for potential future immunotherapeutic inter-
ventions. In fact, it is plausible that virus-​specific or
tumour-​specific CD8+ T cells — either activated with
therapeutic vaccines or other immune stimulators, or
exogenously infused (as in the case of CAR T cells
or TCR-​redirected T cells) — may not be fully functional
in fibrotic/cirrhotic livers.
Targeting immunopathology
Morbidity and mortality in patients with CHB increase
as liver fibrosis, cirrhosis and, especially, HCC develop4.
Hence, therapeutic approaches aimed at delaying or
preventing the onset of these complications may be
beneficial. One option would be to limit T cell-​induced
immunopathology without inducing generalized
immune suppression. The discovery that platelets sup-
port the homing of pathogenic effector CD8+ T cells into
the liver124, and that this process is hindered by specific
platelet inhibitors including aspirin19, prompted the
preclinical evaluation of platelet inhibitors in CHB. In
mouse models of immune-​mediated CHB, treatment
with aspirin and/or other anti-​platelet drugs reduced the
accumulation of virus-​specific T cells within the liver,
as well as liver damage18. As a consequence, fibrosis and
compensatory hepatocellular proliferation were reduced,
HCC development was prevented and overall survival
improved18. Mechanistically, low-​dose aspirin — which
inhibits the enzyme COX1 and has almost exclusively
platelet-​specific effects, with little or no interference
with COX2 expression and/or activity125 — specif-
ically hampers the capacity of platelets to form the
intrasinusoidal small aggregates that function as dock-
ing sites for circulating virus-​specific CD8+ T cells89.
Of note, the antitumour effect of anti-​platelet therapy
could have also resulted from additional inhibition of
platelet-​derived oncogenic products such as serotonin
(stimulating hepatocellular proliferation)126 and numer-
ous growth factors (including epidermal growth factor,
insulin-​like growth factor, platelet-​derived growth factor
and vascular endothelial growth factor). However, the
experimental evidence that platelet inhibition did not
prevent chemical hepatocarcinogenesis18 suggests that
anti-​platelet therapy inhibits HCC development during
CHB mainly by limiting the intrahepatic accumulation
of HBV-​specific T cells.
It is of note that, following these preclinical studies,
a large number of meta-​analyses similarly reported that
patients with CHB who undergo low-​dose aspirin treat-
ment have a reduced incidence of HCC and reduced
mortality, with no significant side effects17,127–134. These
findings should motivate a constructive discussion by
the international scientific community on whether the
preclinical and clinical results collected thus far repre-
sent sufficient background evidence to warrant the use
of low-​dose aspirin in patients with CHB or whether
they should prompt prospective randomized clinical
trials with careful patient selection and monitoring135.
Concluding remarks
Even though much has been learned about the molecular
biology of HBV and the disease correlates of AHB and
CHB, several immunologic and pathogenic aspects that
potentially influence the outcome of infection remain
poorly understood. Novel DAAs are being continuously
developed for the treatment of CHB, and these include mol-
ecules that target the biogenesis and persistence of cccDNA
in the hepatocyte136. Whether these DAAs, alone or in com-
bination with existing or forthcoming antivirals, will be able
to achieve a long‐lasting functional cure is unclear. It is
also unclear whether strategies aimed at boosting immune
responses to HBV or dampening immunopathology will
have a positive impact in this scenario. Undoubtedly, the
development of curative strategies for CHB will greatly
benefit from a deeper understanding of the mechanisms
that govern HBV immunobiology and pathogenesis, which
may guide the design of novel, evidence-​based clinical stud-
ies aimed at terminating persistent HBV infection and its
life-​threatening complications.
Published online xx xx xxxx

1. Guidotti, L. G. & Chisari, F. V. Immunobiology and
pathogenesis of viral hepatitis. Annu. Rev. Pathol.
Mech. Dis. 1, 23–61 (2006).
2. Locarnini, S., Hatzakis, A., Chen, D.-S. & Lok, A.
Strategies to control hepatitis B: public policy,
epidemiology, vaccine and drugs. J. Hepatol. 62,
S76–S86 (2015).
3. Yuen, M.-F. et al. Hepatitis B virus infection. Nat. Rev.
Dis. Primers 4, 18035 (2018).
4. Revill, P. A. et al. A global scientific strategy to cure
hepatitis B. Lancet Gastroenterol. Hepatol. 4,
545–558 (2019).
5. Udompap, P. & Kim, W. R. Development of
hepatocellular carcinoma in patients with suppressed
viral replication: changes in risk over time. Clin. Liver
Dis. 15, 85–90 (2020).
6. Levrero, M., Testoni, B. & Zoulim, F. HBV cure:
why, how, when? Curr. Opin. Virol. 18, 135–143
(2016).
7. Fanning, G. C., Zoulim, F., Hou, J. & Bertoletti, A.
Therapeutic strategies for hepatitis B virus infection:
towards a cure. Nat. Rev. Drug Discov. 18, 827–844
(2019).
8. Rehermann, B., Ferrari, C., Pasquinelli, C.
& Chisari, F. V. The hepatitis B virus persists for
decades after patients’ recovery from acute viral
hepatitis despite active maintenance of a cytotoxic
T-lymphocyte response. Nat. Med. 2, 1104–1108
(1996).
9. Kim, C. Y. & Tilles, J. G. Purification and biophysical
characterization of hepatitis B antigen. J. Clin. Invest.
52, 1176–1186 (1973).
10. Seeger, C. & Mason, W. S. Molecular biology of
hepatitis B virus infection. Virology 479, 672–686
(2015).
11. Bertoletti, A. & Ferrari, C. Adaptive immunity in HBV
infection. J. Hepatol. 64, S71–S83 (2016).
12. Guidotti, L. G., Isogawa, M. & Chisari, F. V. Host–virus
interactions in hepatitis B virus infection. Curr. Opin.
Immunol. 36, 61–66 (2015).
13. Tu, T. et al. Integration occurs early in the viral life
cycle in an in vitro infection model via sodium
taurocholate cotransporting polypeptide-​dependent
uptake of enveloped virus particles. J. Virol. 92,
e02007–e02017 (2018).
This study shows that HBV DNA integration
occurs early upon infection in an in vitro
infection model.

NATuRE REVIEWS | Immunology

0123456789();:
Reviews
14. Summers, J. et al. Hepatocyte turnover during
resolution of a transient hepadnaviral infection. Proc.
Natl Acad. Sci. USA 100, 11652–11659 (2003).
15. Yang, W. & Summers, J. Integration of hepadnavirus
DNA in infected liver: evidence for a linear precursor.
J. Virol. 73, 9710–9717 (1999).
16. Wooddell, C. I. et al. RNAi-​based treatment of
chronically infected patients and chimpanzees reveals
that integrated hepatitis B virus DNA is a source
of HBsAg. Sci. Transl Med. 9, eaan0241 (2017).
This paper reveals integrated HBV DNA as a relevant
source of HBsAg in patients and chimpanzees with
chronic infection.
17. Simon, T. G. et al. Association of aspirin with hepato­
cellular carcinoma and liver-​related mortality. N. Engl.
J. Med. 382, 1018–1028 (2020).
This manuscript represents one of a large number
of meta-​analyses describing an association between
low-​dose aspirin treatment and reduced HCC
incidence.
18. Sitia, G. et al. Antiplatelet therapy prevents
hepatocellular carcinoma and improves survival in a
mouse model of chronic hepatitis B. Proc. Natl Acad.
Sci. USA 109, E2165–E2172 (2012).
This preclinical study shows that anti-​platelet
therapy reduces liver fibrosis and prevents HCC
in mouse models of CHB.
19. Iannacone, M., Sitia, G., Narvaiza, I., Ruggeri, Z. M.
& Guidotti, L. G. Antiplatelet drug therapy moderates
immune-​mediated liver disease and inhibits viral
clearance in mice infected with a replication-​deficient
adenovirus. Clin. Vaccine Immunol. 14, 1532–1535
(2007).
20. Jilbert, A. R., Miller, D. S., Scougall, C. A., Turnbull, H.
& Burrell, C. J. Kinetics of duck hepatitis B virus
infection following low dose virus inoculation: one virus
DNA genome is infectious in neonatal ducks. Virology
226, 338–345 (1996).
21. Asabe, S. et al. The size of the viral inoculum
contributes to the outcome of hepatitis B virus
infection. J. Virol. 83, 9652–9662 (2009).
22. Wisse, E., Jacobs, F., Topal, B., Frederik, P.
& Geest, B. D. The size of endothelial fenestrae in
human liver sinusoids: implications for hepatocyte-​
directed gene transfer. Gene Ther. 15, 1193–1199
(2008).
23. Vollmar, B. & Menger, M. D. The hepatic
microcirculation: mechanistic contributions and
therapeutic targets in liver injury and repair. Physiol.
Rev. 89, 1269–1339 (2009).
24. Whalley, S. A. et al. Kinetics of acute hepatitis B virus
infection in humans. J. Exp. Med. 193, 847–854
(2001).
25. Guidotti, L. G., Matzke, B., Schaller, H. & Chisari, F. V.
High-​level hepatitis B virus replication in transgenic
mice. J. Virol. 69, 6158–6169 (1995).
26. Guidotti, L. G. et al. Viral clearance without destruction
of infected cells during acute HBV infection. Science
284, 825–829 (1999).
27. Wieland, S. F., Spangenberg, H. C., Thimme, R.,
Purcell, R. H. & Chisari, F. V. Expansion and contraction
of the hepatitis B virus transcriptional template in
infected chimpanzees. Proc. Natl Acad. Sci. USA 101,
2129–2134 (2004).
28. Wieland, S., Thimme, R., Purcell, R. H. & Chisari, F. V.
Genomic analysis of the host response to hepatitis B
virus infection. Proc. Natl Acad. Sci. USA 101,
6669–6674 (2004).
29. Suslov, A. et al. Virus does not interfere with innate
immune responses in the human liver. Gastroenterology
154, 1778–1790 (2018).
30. Tsui, L. V., Guidotti, L. G., Ishikawa, T. & Chisari, F. V.
Posttranscriptional clearance of hepatitis B virus
RNA by cytotoxic T lymphocyte-​activated hepatocytes.
Proc. Natl Acad. Sci. USA 92, 12398–12402 (1995).
31. Heise, T., Guidotti, L. G., Cavanaugh, V. J. & Chisari, F. V.
Hepatitis B virus RNA-​binding proteins associated with
cytokine-​induced clearance of viral RNA from the liver
of transgenic mice. J. Virol. 73, 474–481 (1999).
32. Heise, T., Guidotti, L. G. & Chisari, F. V. La autoantigen
specifically recognizes a predicted stem-​loop in
hepatitis B virus RNA. J. Virol. 73, 5767–5776 (1999).
33. McClary, H., Koch, R., Chisari, F. V. & Guidotti, L. G.
Relative sensitivity of hepatitis B virus and other
hepatotropic viruses to the antiviral effects of cytokines.
J. Virol. 74, 2255–2264 (2000).
34. Wieland, S. F., Guidotti, L. G. & Chisari, F. V.
Intrahepatic induction of α/β interferon eliminates viral
RNA-​containing capsids in hepatitis B virus transgenic
mice. J. Virol. 74, 4165–4173 (2000).
35. Kimura, K., Kakimi, K., Wieland, S., Guidotti, L. G. &
Chisari, F. V. Activated intrahepatic antigen-​presenting
cells inhibit hepatitis B virus replication in the liver of
transgenic mice. J. Immunol. 169, 5188–5195 (2002).
36. Vilarinho, S., Ogasawara, K., Nishimura, S., Lanier, L. L.
& Baron, J. L. Blockade of NKG2D on NKT cells
prevents hepatitis and the acute immune response
to hepatitis B virus. Proc. Natl Acad. Sci. USA 104,
18187–18192 (2007).
37. Isogawa, M., Robek, M. D., Furuichi, Y. & Chisari, F. V.
Toll-​like receptor signaling inhibits hepatitis B virus
replication in vivo. J. Virol. 79, 7269–7272 (2005).
38. Suslov, A., Wieland, S. & Menne, S. Modulators of
innate immunity as novel therapeutics for treatment
of chronic hepatitis B. Curr. Opin. Virol. 30, 9–17
(2018).
39. Iwasaki, A. A virological view of innate immune
recognition. Annu. Rev. Microbiol. 66, 177–196
(2012).
40. Webster, G. J. M. et al. Incubation phase of acute
hepatitis B in man: dynamic of cellular immune
mechanisms. Hepatology 32, 1117–1124 (2000).
41. Thimme, R. et al. CD8+ T cells mediate viral clearance
and disease pathogenesis during acute hepatitis B
virus infection. J. Virol. 77, 68–76 (2003).
42. Hoofnagle, J. H., Gerety, R. J. & Barker, L. F.
Antibody to hepatitis-​B-virus core in man. Lancet 302,
869–873 (1973).
43. Maini, M. K. & Burton, A. R. Restoring, releasing or
replacing adaptive immunity in chronic hepatitis B.
Nat. Rev. Gastroenterol. Hepatol. 16, 662–675
(2019).
44. Guidotti, L. G. et al. Cytotoxic T lymphocytes inhibit
hepatitis B virus gene expression by a noncytolytic
mechanism in transgenic mice. Proc. Natl Acad. Sci.
USA 91, 3764–3768 (1994).
45. Guidotti, L. G. et al. Intracellular inactivation of the
hepatitis B virus by cytotoxic T lymphocytes. Immunity
4, 25–36 (1996).
46. Wong, Y. C., Tay, S. S., McCaughan, G. W., Bowen, D. G.
& Bertolino, P. Immune outcomes in the liver: is CD8
T cell fate determined by the environment? J. Hepatol.
63, 1005–1014 (2015).
47. Isogawa et al. CD40 activation rescues antiviral CD8+
T cells from PD-1-mediated exhaustion. PLoS Pathog.
9, e1003490 (2013).
48. Bénéchet, A. P. et al. Dynamics and genomic
landscape of CD8+ T cells undergoing hepatic priming.
Nature 574, 200–205 (2019).
This paper reveals that hepatocellular priming
leads to a T cell dysfunction that is refractory to
checkpoint inhibition but responds to IL-2.
49. Bertolino, P. et al. Death by neglect as a deletional
mechanism of peripheral tolerance. Int. Immunol. 11,
1225–1238 (1999).
50. Pol et al. Effects of interleukin-2 in immunostimulation
and immunosuppression. J. Exp. Med. 217, 2261
(2020).
51. Blattman, J. N. et al. Therapeutic use of IL-2 to
enhance antiviral T-​cell responses in vivo. Nat. Med. 9,
540–547 (2003).
52. West, E. E. et al. PD-​L1 blockade synergizes with IL-2
therapy in reinvigorating exhausted T cells. J. Clin.
Invest. 123, 2604–2615 (2013).
53. Kuipery, A., Gehring, A. J. & Isogawa, M. Mechanisms
of HBV immune evasion. Antivir. Res. 179, 104816
(2020).
54. Kennedy, P. T. F. et al. Preserved T-​cell function in
children and young adults with immune-​tolerant
chronic hepatitis B. Gastroenterology 143, 637–645
(2012).
55. Shimizu, Y., Guidotti, L. G., Fowler, P. & Chisari, F. V.
Dendritic cell immunization breaks cytotoxic
T lymphocyte tolerance in hepatitis B virus transgenic
mice. J. Immunol. 161, 4520–4529 (1998).
56. Kakimi, K., Isogawa, M., Chung, J., Sette, A.
& Chisari, F. V. Immunogenicity and tolerogenicity of
hepatitis B virus structural and nonstructural proteins:
implications for immunotherapy of persistent viral
infections. J. Virol. 76, 8609–8620 (2002).
57. Ishak, K. et al. Histological grading and staging of
chronic hepatitis. J. Hepatol. 22, 696–699 (1995).
58. Fisicaro, P. et al. Targeting mitochondrial dysfunction
can restore antiviral activity of exhausted HBV-​specific
CD8 T cells in chronic hepatitis B. Nat. Med. 23,
327–336 (2017).
This article suggests a central role for reactive
oxygen species in T cell exhaustion during CHB,
thus providing novel potential therapeutic targets.
59. Wieland, S. F. The chimpanzee model for hepatitis B
virus infection. CSH Perspect. Med. 5, a021469
(2015).
60. Chen, M. T. et al. A function of the hepatitis B virus
precore protein is to regulate the immune response to
0123456789();:
the core antigen. Proc. Natl Acad. Sci. USA 101,
14913–14918 (2004).
61. Chen, M. et al. Immune tolerance split between
hepatitis B virus precore and core proteins. J. Virol.
79, 3016–3027 (2005).
62. Tian, Y., Kuo, C., Akbari, O. & Ou, J. J. Maternal-​
derived hepatitis B virus e antigen alters macrophage
function in offspring to drive viral persistence after
vertical transmission. Immunity 44, 1204–1214
(2016).
63. Publicover, J. et al. Age-​dependent hepatic lymphoid
organization directs successful immunity to hepatitis B.
J. Clin. Invest. 123, 3728–3739 (2013).
64. Brunetto, M. R. et al. Wild-​type and e antigen-​minus
hepatitis B viruses and course of chronic hepatitis.
Proc. Natl Acad. Sci. USA 88, 4186–4190 (1991).
65. Rivino, L. et al. Hepatitis B virus-​specific T cells
associate with viral control upon nucleos(t)ide-​
analogue therapy discontinuation. J. Clin. Invest. 128,
668–681 (2018).
66. Schuch, A. et al. Phenotypic and functional differences
of HBV core-​specific versus HBV polymerase-​specific
CD8+ T cells in chronically HBV-​infected patients with
low viral load. Gut 68, 905–915 (2019).
67. Fumagalli, V. et al. Serum HBsAg clearance has minimal
impact on CD8+ T cell responses in mouse models of
HBV infection. J. Exp. Med. 217, e20200298 (2020).
This study shows that circulating HBsAg clearance
does not improve HBV-​specific CD8+ T cell responses.
68. Bert, N. L. et al. Effects of hepatitis B surface antigen
on virus-​specific and global T cells in patients with
chronic hepatitis B virus infection. Gastroenterology
159, 652–664 (2020).
69. Li et al. A potent human neutralizing antibody
Fc-dependently reduces established HBV infections.
eLife 6, e26738 (2017).
70. Zhang, T.-Y. et al. Prolonged suppression of HBV in
mice by a novel antibody that targets a unique epitope
on hepatitis B surface antigen. Gut 65, 658 (2015).
71. Neumann et al. Novel mechanism of antibodies to
hepatitis B virus in blocking viral particle release from
cells. Hepatology 52, 875–885 (2010).
72. Galun, E. et al. Clinical evaluation (phase I) of a
combination of two human monoclonal antibodies to
HBV: safety and antiviral properties. Hepatology 35,
673–679 (2002).
73. Bertoletti, A. et al. Cytotoxic T lymphocyte response
to a wild type hepatitis B virus epitope in patients
chronically infected by variant viruses carrying
substitutions within the epitope. J. Exp. Med. 180,
933–943 (1994).
74. Bertoletti, A. et al. Natural variants of cytotoxic
epitopes are T-​cell receptor antagonists for antiviral
cytotoxic T cells. Nature 369, 407–410 (1994).
75. Maini, M. K. et al. T cell receptor usage of virus-​
specific CD8 cells and recognition of viral mutations
during acute and persistent hepatitis B virus infection.
Eur. J. Immunol. 30, 3067–3078 (2000).
76. Bertoletti, A. & Kennedy, P. T. The immune tolerant
phase of chronic HBV infection: new perspectives on
an old concept. Cell Mol. Immunol. 12, 258–263
(2015).
77. Fisicaro, P. et al. Pathogenetic mechanisms of T cell
dysfunction in chronic HBV infection and related
therapeutic approaches. Front. Immunol. 11, 849
(2020).
78. Burton, A. R. et al. Circulating and intrahepatic
antiviral B cells are defective in hepatitis B. J. Clin.
Invest. 128, 4588–4603 (2018).
79. Salimzadeh, L. et al. PD-1 blockade partially recovers
dysfunctional virus-​specific B cells in chronic hepatitis
B infection. J. Clin. Invest. 128, 4573–4587 (2018).
Together with Burton et al. (2018), this paper
detects and characterizes dysfunctional HBsAg-​
specific B cell responses in patients with chronic
HBV infection.
80. Tian, C. et al. Use of ELISpot assay to study
HBs-specific B cell responses in vaccinated and
HBV infected humans. Emerg. Microbes Infec 7, 16
(2018).
81. Xu, X. et al. Reversal of B-​cell hyperactivation and
functional impairment is associated with HBsAg
seroconversion in chronic hepatitis B patients.
Cell Mol. Immunol. 12, 309–316 (2015).
82. Bert, N. L. et al. Comparative characterization
of B cells specific for HBV nucleocapsid and envelope
proteins in patients with chronic hepatitis B. J. Hepatol.
72, 34–44 (2019).
83. Vanwolleghem, T. et al. Hepatitis B core-​specific
memory B cell responses associate with clinical
parameters in patients with chronic HBV. J. Hepatol.
73, 52–61 (2020).
www.nature.com/nri

84. Milich, D. & McLachlan, A. The nucleocapsid of
hepatitis B virus is both a T-​cell-independent and a
T-cell-dependent antigen. Science 234, 1398–1401
(1986).
85. Guidotti, L. G. & Iannacone, M. Effector CD8 T cell
trafficking within the liver. Mol. Immunol. 55, 94–99
(2013).
86. Iannacone, M. Hepatic effector CD8+ T-​cell dynamics.
Cell Mol. Immunol. 12, 269–272 (2015).
87. Inverso, D. & Iannacone, M. Spatiotemporal dynamics
of effector CD8+ T cell responses within the liver.
J. Leukoc. Biol. 99, 51–55 (2016).
88. Benechet, A. P. & Iannacone, M. Determinants of
hepatic effector CD8+ T cell dynamics. J. Hepatol. 66,
228–233 (2017).
89. Guidotti, L. G. et al. Immunosurveillance of the liver by
intravascular effector CD8+ T cells. Cell 161, 486–500
(2015).
This manuscript reports that effector CD8+ T cells
can recognize and kill antigen-​expressing
hepatocytes without extravasating by extending
cytoplasmic protrusions through endothelial
fenestration.
90. Sironi, L. et al. In vivo flow mapping in complex vessel
networks by single image correlation. Sci. Rep. 4,
7341 (2014).
91. Warren, A. et al. T lymphocytes interact with
hepatocytes through fenestrations in murine
liver sinusoidal endothelial cells. Hepatology 44,
1182–1190 (2006).
92. Guidotti, L. G. The role of cytotoxic T cells and
cytokines in the control of hepatitis B virus infection.
Vaccine 20, A80–A82 (2002).
93. Fioravanti, J. et al. Effector CD8+ T cell-​derived
interleukin-10 enhances acute liver immunopathology.
J. Hepatol. 67, 543–548 (2017).
94. Iannacone, M. & Guidotti, L. G. Mouse models of
hepatitis B virus pathogenesis. CSH Perspect. Med. 5,
a021477 (2015).
95. Guidotti, L. G., McClary, H., Loudis, J. M. & Chisari, F. V.
Nitric oxide inhibits hepatitis b virus replication in
the livers of transgenic mice. J. Exp. Med. 191,
1247–1252 (2000).
96. Wieland, S. F., Eustaquio, A., Whitten-​Bauer, C.,
Boyd, B. & Chisari, F. V. Interferon prevents formation
of replication-​competent hepatitis B virus RNA-​
containing nucleocapsids. Proc. Natl Acad. Sci. USA
102, 9913–9917 (2005).
97. Robek, M. D., Wieland, S. F. & Chisari, F. V. Inhibition
of hepatitis B virus replication by interferon requires
proteasome activity. J. Virol. 76, 3570–3574 (2002).
98. Xia, Y. et al. Interferon-​γ and tumor necrosis factor-​α
produced by T cells reduce the HBV persistence form,
cccDNA, without cytolysis. Gastroenterology 150,
194–205 (2016).
99. Michalak, T. I., Pasquinelli, C., Guilhot, S. & Chisari, F. V.
Hepatitis B virus persistence after recovery from acute
viral hepatitis. J. Clin. Invest. 93, 230–239 (1994).
100. Pallett, L. J. et al. IL-2high tissue-​resident T cells in
the human liver: sentinels for hepatotropic infection.
J. Exp. Med. 214, 1567–1580 (2017).
This paper characterizes tissue-​resident memory
T cells in the liver of patients chronically infected
by HBV.
101. Ando, K. et al. Class I-​restricted cytotoxic T
lymphocytes are directly cytopathic for their target
cells in vivo. J. Immunol. 152, 3245–3253 (1994).
102. Nakamoto, Y., Guidotti, L. G., Pasquetto, V.,
Schreiber, R. D. & Chisari, F. V. Differential target
cell sensitivity to CTL-​activated death pathways in
hepatitis B virus transgenic mice. J. Immunol. 158,
5692–5697 (1997).
103. Sitia, G. et al. Kupffer cells hasten resolution of liver
immunopathology in mouse models of viral hepatitis.
PLoS Pathog. 7, e1002061 (2011).
104. Sitia et al. Treatment with HMGB1 inhibitors diminishes
CTL-​induced liver disease in HBV transgenic mice.
J. Leukoc. Biol. 81, 100–107 (2007).
105. Sitia, G. et al. Depletion of neutrophils blocks the
recruitment of antigen-​nonspecific cells into the liver
without affecting the antiviral activity of hepatitis B
virus-​specific cytotoxic T lymphocytes. Proc. Natl
Acad. Sci. USA 99, 13717–13722 (2002).
106. Sitia, G. et al. MMPs are required for recruitment
of antigen-​nonspecific mononuclear cells into the liver
by CTLs. J. Clin. Invest. 113, 1158–1167 (2004).
107. Kakimi, K. et al. Blocking chemokine responsive to
γ-2/interferon (IFN)-γ inducible protein and monokine
induced by IFN-​γ activity in vivo reduces the
pathogenetic but not the antiviral potential of
hepatitis B virus-​specific cytotoxic T lymphocytes.
J. Exp. Med. 194, 1755–1766 (2001).
NATuRE REVIEWS | Immunology
108. Maini, M. K. et al. The role of virus-​specific CD8+
cells in liver damage and viral control during
persistent hepatitis B virus infection. J. Exp. Med.
191, 1269–1280 (2000).
109. Reignat, S. et al. Escaping high viral load exhaustion
CD8 cells with altered tetramer binding in chronic
hepatitis B virus infection. J. Exp. Med. 195,
1089–1101 (2002).
110. Webster, G. J. M. et al. Longitudinal analysis of CD8+
T cells specific for structural and nonstructural
hepatitis B virus proteins in patients with chronic
hepatitis B: implications for immunotherapy. J. Virol.
78, 5707–5719 (2004).
111. Boni, C. et al. Characterization of hepatitis B virus
(HBV)-specific T-​cell dysfunction in chronic HBV
infection. J. Virol. 81, 4215–4225 (2007).
112. Hoogeveen, R. C. et al. Phenotype and function of
HBV-​specific T cells is determined by the targeted
epitope in addition to the stage of infection. Gut 68,
893–904 (2018).
113. Nakamoto, Y., Guidotti, L. G., Kuhlen, C. V., Fowler, P.
& Chisari, F. V. Immune pathogenesis of hepatocellular
carcinoma. J. Exp. Med. 188, 341–350 (1998).
114. Isogawa, M., Furuichi, Y. & Chisari, F. V. Oscillating
CD8+ T cell effector functions after antigen recognition
in the liver. Immunity 23, 53–63 (2005).
115. Khakpoor, A. et al. Spatiotemporal differences in
presentation of CD8 T cell epitopes during hepatitis B
virus infection. J. Virol. 93, e01457-18 (2018).
116. Nakamoto, Y., Suda, T., Momoi, T. & Kaneko, S.
Different procarcinogenic potentials of lymphocyte
subsets in a transgenic mouse model of chronic
hepatitis B. Cancer Res. 64, 3326–3333 (2004).
117. Tang, L. S. Y., Covert, E., Wilson, E. & Kottilil, S.
Chronic hepatitis B infection: a review. JAMA 319,
1802–1813 (2018).
118. Buendia, M.-A. & Neuveut, C. Hepatocellular carcinoma.
CSH Perspect. Med. 5, a021444 (2015).
119. Levrero, M. & Zucman-​Rossi, J. Mechanisms of
HBV-induced hepatocellular carcinoma. J. Hepatol.
64, S84–S101 (2016).
120. Bisceglie, A. M. D. Hepatitis B and hepatocellular
carcinoma. Hepatology 49, S56–S60 (2009).
121. Schuppan, D. & Afdhal, N. H. Liver cirrhosis. Lancet
371, 838–851 (2008).
122. Bataller, R. & Brenner, D. A. Liver fibrosis. J. Clin. Invest.
115, 209–218 (2005).
123. Friedman, S. L. Mechanisms of disease: mechanisms
of hepatic fibrosis and therapeutic implications.
Nat. Clin. Pract. Gastr 1, 98–105 (2004).
124. Iannacone, M. et al. Platelets mediate cytotoxic
T lymphocyte-​induced liver damage. Nat. Med. 11,
1167–1169 (2005).
This study establishes platelets as critical mediators
of liver damage through their capacity to promote
liver homing of effector CD8+ T cells.
125. Ornelas, A. et al. Beyond COX-1: the effects of aspirin
on platelet biology and potential mechanisms of
chemoprevention. Cancer Metast Rev. 36, 289–303
(2017).
126. Haemmerle, M., Stone, R. L., Menter, D. G.,
Afshar-Kharghan, V. & Sood, A. K. The platelet lifeline
to cancer: challenges and opportunities. Cancer Cell
33, 965–983 (2018).
127. Lee, P.-C. et al. Antiplatelet therapy is associated with
a better prognosis for patients with hepatitis B virus-​
related hepatocellular carcinoma after liver resection.
Ann. Surg. Oncol. 23, 874–883 (2016).
128. Hwang, I. C., Chang, J., Kim, K. & Park, S. M. Aspirin
use and risk of hepatocellular carcinoma in a national
cohort study of Korean adults. Sci. Rep. 8, 4968
(2018).
129. Simon, T. G. et al. Association between aspirin use and
risk of hepatocellular carcinoma. JAMA Oncol. 4,
1683 (2018).
130. Lee, T.-Y. et al. Association of daily aspirin therapy with
risk of hepatocellular carcinoma in patients with chronic
hepatitis B. JAMA Intern. Med. 179, 633–640 (2019).
131. Wang, S. et al. Association of aspirin therapy with
risk of hepatocellular carcinoma: a systematic review
and dose–response analysis of cohort studies with
2.5 million participants. Pharmacol. Res. 151,
104585 (2019).
132. Liao, Y.-H. et al. Aspirin decreases hepatocellular
carcinoma risk in hepatitis C virus carriers: a
nationwide cohort study. BMC Gastroenterol. 20, 6
(2020).
133. Bosetti, C., Santucci, C., Gallus, S., Martinetti, M.
& Vecchia, C. L. Aspirin and the risk of colorectal
and other digestive tract cancers: an updated meta-​
analysis through 2019. Ann. Oncol. 31, 558–568
(2020).
0123456789();:
Reviews
134. Hayashi, T. et al. Antiplatelet therapy improves the
prognosis of patients with hepatocellular carcinoma.
Cancers 12, 3215 (2020).
135. Guidotti, L. G., Vecchia, C. L. & Colombo, M. Is it
time to recommend low-​dose aspirin treatment
for the prevention of hepatocellular carcinoma?
Gastroenterology 159, 1988–1990 (2020).
136. Martinez, M. G., Villeret, F., Testoni, B. & Zoulim, F.
Can we cure hepatitis B virus with novel direct-​acting
antivirals? Liver Int. 40, 27–34 (2020).
137. Hillis, W. D. Viral hepatitis associated with sub-​human
primates. Transfusion 3, 445–454 (1963).
138. Walter, E., Keist, R., Niederöst, B., Pult, I. & Blum, H. E.
Hepatitis B virus infection of tupaia hepatocytes in vitro
and in vivo. Hepatology 24, 1–5 (1996).
139. Schulze, A., Gripon, P. & Urban, S. Hepatitis B virus
infection initiates with a large surface protein-​dependent
binding to heparan sulfate proteoglycans. Hepatology
46, 1759–1768 (2007).
140. Sureau, C. & Salisse, J. A conformational heparan
sulfate binding site essential to infectivity overlaps
with the conserved hepatitis B virus A-​determinant.
Hepatology 57, 985–994 (2013).
141. Roskams, T. et al. Heparan sulfate proteoglycan
expression in normal human liver. Hepatology 21,
950–958 (1995).
142. Yan, H. et al. Sodium taurocholate cotransporting
polypeptide is a functional receptor for human
hepatitis B and D virus. eLife 1, e00049 (2012).
143. Döring, B., Lütteke, T., Geyer, J. & Petzinger, E. The
SLC10 carrier family: transport functions and molecular
structure. Curr. Top. Membr. 70, 105–168 (2012).
144. Hu, J. & Liu, K. Complete and incomplete hepatitis B
virus particles: formation, function, and application.
Viruses 9, 56 (2017).
145. Seitz, S., Habjanič, J., Schütz, A. K. & Bartenschlager, R.
The hepatitis B virus envelope proteins: molecular
gymnastics throughout the viral life cycle. Ann. Rev.
Virol. 7, 1–26 (2020).
146. Wisse, E., de Zanger, R. B., Charels, K.,
Van Der Smissen, P. & McCuskey, R. S. The liver sieve:
considerations concerning the structure and function
of endothelial fenestrae, the sinusoidal wall and the
space of disse. Hepatology 5, 683–692 (1985).
147. Ficht, X. & Iannacone, M. Immune surveillance
of the liver by T cells. Sci. Immunol. 5, eaba2351
(2020).
148. Iwakiri, Y. The lymphatic system: a new frontier in
hepatology. Hepatology 64, 706–707 (2016).
149. Jenne, C. N. & Kubes, P. Immune surveillance by the
liver. Nat. Immunol. 14, 996–1006 (2013).
150. Horst, A. K., Neumann, K., Diehl, L. & Tiegs, G.
Modulation of liver tolerance by conventional
and nonconventional antigen-​presenting cells and
regulatory immune cells. Cell Mol. Immunol. 13,
277–292 (2016).
151. Wong, Y. C., McCaughan, G. W., Bowen, D. G. &
Bertolino, P. The CD8 T-​cell response during tolerance
induction in liver transplantation. Clin. Transl Immunol.
5, e102 (2016).
152. Mason, W. S. et al. HBV DNA integration and clonal
hepatocyte expansion in chronic hepatitis B patients
considered immune tolerant. Gastroenterology 151,
986–998.e4 (2016).
153. Tu, T., Budzinska, M. A., Shackel, N. A. & Urban, S.
HBV DNA integration: molecular mechanisms and
clinical implications. Viruses 9, 75 (2017).
154. Budzinska, M. A., Shackel, N. A., Urban, S. & Tu, T.
Cellular genomic sites of hepatitis B virus DNA
integration. Genes 9, 365 (2018).
155. Huang, Z. M. & Yen, T. S. Dysregulated surface gene
expression from disrupted hepatitis B virus genomes.
J. Virol. 67, 7032–7040 (1993).
156. Dienes, H. P. et al. Hepatic expression patterns
of the large and middle hepatitis B virus surface
proteins in viremic and nonviremic chronic hepatitis B.
Gastroenterology 98, 1017–1023 (1990).
157. Chisari, F. V. et al. Molecular pathogenesis of
hepatocellular carcinoma in hepatitis B virus
transgenic mice. Cell 59, 1145–1156 (1989).
158. Su, I., Wang, H., Wu, H. & Huang, W. Ground glass
hepatocytes contain pre-​S mutants and represent
preneoplastic lesions in chronic hepatitis B virus
infection. J. Gastroen Hepatol. 23, 1169–1174 (2008).
159. Hadziyannis, S., Gerber, M. A., Vissoulis, C.
& Popper, H. Cytoplasmic hepatitis B antigen in
“ground-​glass” hepatocytes of carriers. Arch. Pathol.
96, 327–330 (1973).
160. Tu, T. et al. Clonal expansion of hepatocytes with
a selective advantage occurs during all stages of
chronic hepatitis B virus infection. J. Viral Hepat. 22,
737–753 (2015).
Reviews
Acknowledgements
The authors thank M. Silva for secretarial assistance,
F. Andreata for help with figure preparation and the members of
the Iannacone and Guidotti laboratories for helpful discussions.
They apologize to all authors whose work they could not cite due
to space constraints. M.I. is supported by the European
Research Council (ERC) Consolidator Grant 725038, ERC Proof
of Concept Grant 957502, Italian Association for Cancer
Research (AIRC) Grants 19891 and 22737, Italian Ministry of
Health (MoH) Grants RF-2018-12365801 and COVID-2020-
12371617, Lombardy Foundation for Biomedical Research
(FRRB) Grant 2015-0010, the European Molecular Biology
Organization Young Investigator Program and a Funded
Research Agreement from Gilead Sciences. L.G.G. is supported
by the AIRC Grant 22737, Lombardy Open Innovation Grant
229452, PRIN Grant 2017MPCWPY from the Italian Ministry
of Education, University and Research, and Funded Research
Agreements from Gilead Sciences, Avalia Therapeutics and
CNCCS SCARL.
Author contributions
M.I. and L.G.G. contributed equally to this work.
Competing interests
M.I. participates in advisory boards/consultancies for Gilead
Sciences, Roche, Third Rock Ventures, Amgen, Asher Bio and
Allovir. L.G.G is a member of the board of directors at
Genenta Science and Epsilon Bio and participates in advisory
boards/consultancies for Gilead Sciences, Roche and Arbutus
Biopharma. M.I. and L.G.G. are inventors on patents filed,
owned and managed by San Raffaele Scientific Institute,
Vita-​Salute San Raffaele University and Telethon Foundation
0123456789();:
on technology related to work discussed in this manuscript
( WO 2 0 2 0 / 01 6 4 3 4 , WO 2 0 2 0 / 01 6 4 2 7 , WO 2 0 2 0 /
030781, WO2020/234483, EU patent applications n.
19211249.8 and n. 20156716.1, and UK patent application n.
1907493.9).
Peer review information
Nature Reviews Immunology thanks Anna Lok, Antonio
Bertoletti and Mala Maini for their contribution to the peer
review of this work.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
© Springer Nature Limited 2021
www.nature.com/nri