 |
|
Laboratory of Experimental Virology
Research Projects
The HBV Specific Cytotoxic T Lymphocyte
Response in Infected Patients
The peripheral blood cytotoxic T
lymphocyte (CTL) response to the hepatitis B virus (HBV) is vigorous, polyclonal
and multispecific in patients with acute viral hepatitis (AVH), it persists
indefinitely after recovery, and it is maintained by continued antigenic
stimulation by residual virus that persists, apparently harmlessly, in healthy
convalescent individuals. In contrast, the CTL response to HBV is relatively
weak in patients with chronic hepatitis, except during spontaneous disease
flares or interferon induced recovery when it is readily detectable. Collectively,
these results suggest that the CTL response contributes both to viral clearance
and to disease pathogenesis in HBV infection when it is strong, and that
it contributes to viral persistence and chronic liver disease when it is
weak.
Based on these results, therapeutic
induction or augmentation of the CTL response to HBV may lead to viral clearance
in chronically infected patients. Experimental CTL vaccines must include
several CTL epitopes for each class I allele in order to avoid the selection
of CTL escape mutants. We have recently identified several new HBV-specific
CTL epitopes that are restricted by multiple class I alleles in the HLA-A2,
HLA-A3 and HLA-B7 superfamilies, including a new HLA B51-restricted epitope
that is located between HBc residues 19-27. This epitope is particularly
interesting because it is contained entirely within a previously described
HLA A2-restricted CTL epitope located between HBc residues 18-27. Interestingly,
HLA-B57 restricted CTLs specific for HBc19-27 can kill HLA-B51 positive
target cells pulsed with the HBc18-27, implying that it can be presented
by HLA-B51 as well as by HLA A2 and suggesting that, because of its broad
reactivity, the longer peptide should be included in the design of CTL-based
vaccines.
The HCV Specific T Lymphocyte
Response in Infected Patients
In contrast to HBV infection,
the CTL response is relatively strong during chronic HCV infection yet the
virus persists, suggesting that the host-virus interactions that determine
the outcome of these two infections may be different for each virus. Recently,
we showed that the HCV-specific CD4+ T cell response is stronger
in convalescent than in chronically infected individuals and that this response
is maintained indefinitely after viral clearance. In contrast, HCV-specific
CD8+ CTL activity is higher during chronic infection than after
viral clearance, although CD8+ T cell lines from chronically
infected patients are less likely to produce interferon gamma (IFNg)
than those derived from convalescent individuals. These results suggest
that the CD4+ T cell response to HCV is more closely associated
with HCV clearance than the CD8+ T cell response which is more
often associated with viral persistence and chronic liver disease, perhaps
because its cytolytic activity is not balanced by its ability to produce
IFNg.
In order to define the characteristics
of the antiviral T cell response during the early stages of infection, the
proliferative T cell response to HCV was studied in subjects who experienced
high risk needlestick exposures to HCV. Patients who became infected developed
strong proliferative T cell responses to HCV proteins beginning between
4-8 weeks after exposure, corresponding with anti-HCV seroconversion and
with the clinical onset of viral hepatitis. Despite these T cell responses,
both of these patients became persistently infected. These results demonstrate
that the proliferative T cell response to HCV is a relatively early event
during HCV infection, and that it coincides with the appearance of antibodies
and transaminase elevations but it does not lead to viral clearance.
The T cell Response to HCV
in Acutely Infected Chimpanzees
The early events in immune recognition
during HCV infection are not well understood; nor is the role of T cell
priming in resistance to HCV reinfection. To define these events, we are
studying the peripheral and intrahepatic T cell response to HCV in chimpanzees
following either intrahepatic transfection of cloned HCV RNA or intravenous
infection with monoclonal virus. Our observations thus far can be summarized
as follows. First, HCV specific CD4+ T cells appear in the peripheral
blood as early as 2 weeks after infection, i.e. soon after the appearance
of viral RNA and several weeks before the appearance of HCV-specific CD8+
T cells and antibodies. Second, the HCV-specific T cells appear
in the liver after they are detectable in the peripheral blood. Third, the
T cell responses have not been associated with any changes in serum HCV
RNA titers and they are usually not associated with elevated alanine aminotransferase
activity. Fourth, prior infection or exposure to HCV does not protect against
infection with homologous or heterologous HCV inocula despite the presence
of HCV-specific T cells in the peripheral blood. Fifth, the T cell response
to HCV in chronically infected chimpanzees can be as vigorous and multispecific
as in animals that clear the virus. These results suggest that the T cell
response to HCV may not control viral replication during HCV infection.
In addition, the data provide the first evidence that, like the antibody
response, T cell priming by prior exposure to HCV does not prevent infection
by homologous virus.
Host-Virus Interactions in
HBV Transgenic Mice and HBV-infected Chimpanzees
Class I restricted HBV-specific
murine CTLs cause a necroinflammatory liver disease in HBV transgenic mice
that express all of the viral gene products and replicate the virus in their
hepatocytes. In addition to killing the hepatocytes, the CTLs inhibit HBV
gene expression and replication in viable hepatocytes by a noncytolytic
process that is mediated by IFNg and tumor necrosis
factor alpha (TNFa). Importantly, the same noncytolytic
antiviral effects can be triggered by the adoptive transfer of HBV-specific,
class II restricted CD4+ T cells, as well as by HBV-nonspecific
stimuli such as the injection of recombinant IL-12 or during lymphocytic
choriomeningitis virus, murine cytomegalovirus and recombinant adenovirus
infection. The antiviral effects of these agents are also due to their ability
to induce IFNg, TNFa and type I
interferon (IFNa/b) in the liver. These observations
imply that HBV may be susceptible to similar regulatory processes during
natural infection. This hypothesis is supported by our recent observation
that CD4+ and CD8+ T cells and
cytokine markers appear in the liver and that viral DNA disappears from
the liver and serum prior to the onset of acute hepatitis in chimpanzees
inoculated with serum from the HBV transgenic mice. Collectively, these
observations suggest that viral clearance during HBV and, perhaps, other
viral infections may be mediated by noncytolytic antiviral signals triggered
by inflammatory cytokines including IFNg, TNFa
and type 1 IFN in addition to the conventional concept of immune-mediated
destruction of infected cells.
We do not know, however, if these
cytokines represent the final mediators in this system or if other downstream
factors are involved. Nitric oxide (NO) is a highly unstable product of
L-arginine metabolism that exerts a wide variety of biological functions
including antiviral activities in a paracrine and autocrine fashion. NO
is produced by two distinct categories of nitric oxide synthase isoforms.
The constitutive forms are present in many different cell types (including
endothelial and neuronal cells) that synthesize small amounts of NO. The
inducible forms of nitric oxide synthase (iNOS) are present in macrophages
and hepatocytes that synthesize large amounts of NO upon activation by cytokines
such as IFNg and TNFa. To examine
the ability of NO to control HBV replication, we crossed HBV transgenic
mice that replicate the virus in their hepatocytes with mice that lack iNOS.
Interestingly, iNOS knockout mice replicate HBV in the liver at higher levels
than their littermate controls that express iNOS, suggesting that NO present
in the normal liver may repress HBV replication. More importantly, HBV-specific
CTLs did not inhibit HBV replication in iNOS-deficient mice, despite the
fact that they caused a necroinflammatory liver disease and induced IFNg
and TNFa in the liver. In contrast, iNOS-deficient
mice were not resistant to the antiviral effects of the IFNa/b-inducer
polyIC on HBV replication. These results suggest that NO mediates the ability
of IFNg and/or TNFa to inhibit HBV
replication while the antiviral activity of IFNa/b is
NO-independent.
Inhibition of HBV Replication
by Activated Macrophages During Murine Malaria
In keeping with the foregoing
observations, we speculated that additional HBV-nonspecific proinflammatory
events occurring in the liver would inhibit HBV replication. Malaria infection
affects the liver and it is widespread in areas where HBV is endemic. We
have recently shown that intrahepatic HBV replication is inhibited in mice
infected by plasmodium species that cause murine malaria. When injected
into susceptible mice, malaria sporozoites infect and multiply in hepatocytes
where they develop into merozoites and are released into the circulation.
The merozoites are not infectious for hepatocytes; instead, they infect
and multiply in erythrocytes that lyse and release new merozoites to establish
a massive erythrocytic infection that is ultimately controlled by the immune
response. Lysed erythrocytes, hemoglobin and malaria pigment are removed
from the blood by splenic and hepatic macrophages (Kupffer cells), resulting
in macrophage activation, hyperplasia and the recruitment of inflammatory
cells into these organs. In the current study, HBV transgenic mice were
infected either with plasmodium sporozoites or merozoite-infected erythrocytes.
The animals developed Kupffer cell hyperplasia and a diffuse T cell infiltrate
coinciding with parasitemia, the induction of IFNa/b
TNFa and IFNg mRNA and the disappearance
of HBcAg and HBV replicative intermediates from the liver. Importantly,
infection of hepatocytes was not required for this effect since HBV replication
was inhibited efficiently when the infection was limited to erythrocytes.
These results strongly suggest that, during murine malaria, HBV replication
is inhibited by inflammatory cytokines secreted primarily by intrahepatic
macrophages that are activated by phagocytosis of merozoite-infected erythrocytes.
Molecular Basis for Inhibition
of HBV Replication by Inflammatory Cytokines
Several studies are underway
to define the intracellular molecular basis for the antiviral effects of
IFNg and TNFa on HBV gene expression
and replication. One study focuses on the mechanism whereby the cytokines
eliminate cytoplasmic HBV capsids and replicative intermediates from the
cell. To determine whether this is due to translational inhibition of HBV
mRNA we examined the translational status of HBV transcripts in the liver
before and after HBV replication had been blocked by inflammatory cytokines.
The polysomal distribution of HBV specific RNAs did not change after administration
of IL-12 or the IFN type I inducer poly I/C, despite the disappearance of
capsids and replicative intermediates, suggesting that capsid elimination
is a post-translational event. Analysis of the kinetics of elimination of
encapsidated RNA and DNA from the livers of poly I/C-treated mice revealed
that RNA-containing capsids were eliminated before capsids containing immature
HBV DNA, and the immature DNA-containing capsids disappeared before capsids
containing mature HBV DNA. These results suggest that the cytokines inhibit
an early step in HBV replication between translation and capsid maturation
rather than by degrading preformed particles, inhibiting capsid maturation
or accelerating the export of capsids from the cell. This was supported
in BrdU-labeling experiments demonstrating that the kinetics of HBV replication
and release of virions are unaffected when cytoplasmic capsids and HBV replicative
intermediates are disappearing from the cell. It was further supported by
demonstrating that the kinetics of HBV clearance from the serum of poly
I/C-treated mice parallels the half life of virions in the serum of untreated
mice. Collectively, these results suggest that inflammatory cytokines inhibit
encapsidation of HBV pgRNA while the preformed viral capsids and replicative
intermediates are depleted from the cell by normal viral maturation and
secretion.
We have also demonstrated that
the cytokines eliminate HBV RNA by a post-transcriptional mechanism. While
examining the molecular basis of this process, we recently showed that the
cellular La protein and its cleavage products bind a predicted RNA stem
loop structure between nts 1275 and 1291 of HBV RNA. The presence of the
full-length La protein in liver nuclear extracts is associated with normal
levels of HBV RNA in the liver of HBV transgenic mice. In contrast, cytokine-induced
suppression of HBV RNA content is associated with the proteolytic cleavage
of La. Recently, we identified endonucleolytic cleavage sites within the
viral RNA close to the La-binding site. Preliminary results indicate that
recombinant La can reduce the cleavage of HBV RNA by cellular RNases. Taken
together, these results suggest that hepatocellular HBV RNA is stabilized
by full-length La by preventing RNA cleavage and that cytokine-mediated
proteolytic processing of La abrogates this activity and destabilizes HBV
RNA by exposing specific endoribonucleolytic cleavage sites close to the
La binding site.
Thymic Tolerance to One Viral
Protein Reduces Viral-Induced Immunopathology
Under certain circumstances,
cytokine activation in virus-infected tissues can also be harmful and it
can lead to immunopathology, as occurs during LCMV infection of perforin-deficient
mice. Perforin-deficient mice do not clear LCMV infection and they usually
die within few weeks of the infection. In these animals, death is associated
with severe immunopathology in various organs in which high numbers of viral-specific
T cells are constantly activated and produce inflammatory cytokines. We
have recently tested in this model whether it is possible to specifically
dampen the immune response and therefore reduce the immunopathology and
prolong the survival of the animals. To do so, we crossed perforin-deficient
mice with LCMV-transgenic mice that are immunologically tolerant to a single
LCMV antigen that is expressed in the thymus. Upon LCMV infection, the perforin-deficient
LCMV-transgenic mice show reduced immunopathology and in most cases survive
the infection for several months, thus indicating that sometimes it can
be beneficiary for the host to specifically dampen the immune response.
TSRI > IMS > Exp Vir
|
 |