Intracellular Hepatitis B Virus Nucleocapsids Survive Cytotoxic T-Lymphocyte-Induced Apoptosis^†

HBsAg-specific CTLs induce apoptosis in primary hepatocytes derived from HBV-transgenic mice.

First, we analyzed whether HBsAg-specific CTLs could induce apoptosis in hepatocytes that replicate HBV. The HBV-transgenic mice used in this study, lineage 1.3.46 (official designation, Tg [HBV 1.3 genome] Chi46), replicate high levels of HBV in the liver without any evidence of cytopathology (13). For these experiments, age (8 to 10 weeks)-, sex (female)-, and hepatitis B e antigen (HBeAg)-matched [1.3.46 (B10D2) × BALB/c]F₁ mice were used. HBeAg levels in their sera were measured using a commercially available kit from Abbott Laboratories, Abbott Park, Ill. Primary hepatocytes, isolated from these mice by collagenase digestion as described (9, 17), were incubated with a previously described HBsAg-specific CTL clone, 6C2, which is Ld restricted and recognizes an HBsAg epitope (IPQSLDSWWTSL) derived from residues 28 to 39 of the HBV small-envelope protein (1, 14).

The HBsAg-specific CTLs were incubated with the hepatocytes at an effector/target (E/T) ratio of 1:1 for 10, 60, 120, and 240 min, in the presence or absence of 100 μM caspase-3 inhibitor II (Z-DEVD-FMK) (Calbiochem), a potent, cell-permeable, and irreversible inhibitor of CPP-32/apopain, a member of the ICE/CED-3 family of caspases (21). Genomic DNA isolated from a pool of the adherent cells, plus any detached cells and apoptotic debris present in the culture supernatant, was then analyzed for the presence of DNA fragmentation (28). DNA was extracted by lysis with 0.5% sodium dodecyl sulfate and proteinase K digestion (500 μg/ml), followed by phenol extraction and ethanol precipitation. Ten micrograms of this DNA was electrophoresed through 1.5% agarose gels containing 5 μg of ethidium bromide per ml and examined on a UV transilluminator. In all experiments, cell death was also monitored by measuring alanine aminotransferase (ALT) activity in the supernatants in a Paramax chemical analyzer (Baxter Diagnostics Inc., McGaw Park, Ill.).

As shown in Fig. 1, DNA ladder formation, a marker for chromosomal DNA fragmentation, was detected as early as 60 min after coincubation of CTLs and primary hepatocytes. It became maximal between 1 and 2 h and did not increase thereafter. Similarly, hepatocellular lysis, detected by the release of ALT into the culture supernatant (Fig. 1), was also detectable at 60 min but increased for up to 4 h after the addition of CTLs, suggesting that the apoptotic hepatocytes eventually disintegrated and released their cytosolic contents into the media. As shown in Fig. 1, DNA ladder formation was not observed when the primary hepatocytes were preincubated with Z-DEVD-FMK, indicating that CTL-dependent cell death was apoptotic. In order to confirm that DNA ladder formation reflected apoptosis of hepatocytes rather than activation-induced cell death of the CTLs, nonadherent cells from the CTL-hepatocyte cocultures were stained for CD8 expression with fluorescein-isothiocyanate-labeled anti-CD8 antibody (Pharmingen, San Diego, Calif.), and 7-amino-actinomycin D (7-AAD) (Molecular Probes, Eugene, Oreg.) was also added for detection of apoptosis and dead-cell exclusion, as described (15, 23). Multivariate analysis of the data was performed with CELLQuest software (Becton Dickinson Immunocytometry Systems). As indicated at the bottom of Fig. 1, the percentage of apoptotic (7-AAD-positive) cells in the CD8-positive population was 6.1% after 10 min of coincubation with the hepatocytes, at which time point no DNA ladder formation was detectable, and it did not change throughout the experiment (Fig. 1, bottom) despite the progression of DNA ladder formation, indicating that the CTLs did not contribute to the DNA ladder formation observed in this study.

Intracellular HBV nucleocapsids are resistant to hepatocyte apoptosis.

Next, we analyzed whether intracellular HBV nucleocapsids were destroyed by the apoptotic process. Adherent cells, nonadherent cells and cell debris, and cell culture supernatants were examined for the presence of HBV DNA-containing nucleocapsids before and after incubation with CTLs by native agarose-gel electrophoresis of intact nucleocapsids followed by Western and Southern blot analyses (4). Intact, viable adherent cells were lysed with 0.5 ml of NP-40 buffer (50 mM Tris-HCl [pH 7.4], 1 mM EDTA, 0.2% Nonidet P-40) in the culture dish, and a crude cytoplasmic extract was obtained after pelleting the nuclei by centrifugation. Nonadherent cells and cell debris lysed in the same buffer were pelleted from the culture supernatant by a 5-min centrifugation at 500 × g. HBV nucleocapsids in the remaining cleared culture supernatant were isolated by pelleting through a 20% sucrose cushion as described (16) and lysed as described above. Ten microliters of pooled cellular extracts or of extracts isolated from the cleared supernatants was subjected to nondenaturing agarose-gel electrophoresis as described (4). After electrophoresis, proteins were transferred onto nitrocellulose membranes by capillary blotting in TNE (1 mM Tris-HCl [pH 7.4], 1 mM EDTA, 150 mM NaCl) and the membranes were washed for 10 min in water and air dried. HBV nucleocapsids were detected by Western blotting with a rabbit anti-core antibody (Dako, Carpinteria, Calif.) followed by peroxidase-labeled goat anti-rabbit immunoglobulin G (Sigma Biochemicals, St. Louis, Mo.) and chemiluminescence (Pierce, Rockford, Ill.). As expected, in untreated primary hepatocytes most of the HBV capsid antigen (HBcAg) was found in the cell fraction (Fig. 2A). After CTL-induced hepatocellular apoptosis, however, the majority of HBcAg was detectable in the cell-free supernatant (Fig. 2A). After Western blotting, the encapsidated nucleic acids were released by alkali treatment and detected by Southern blotting, as described (4). As shown in Fig. 2A, encapsidated HBV DNA that was detected in the untreated living hepatocytes was released into the culture supernatant after CTL-induced apoptosis.

These results suggest that the HBcAg- and HBV-replicative DNA intermediates are resistant to the CTL-induced enzymatic degradation of the host cell, implying that the nucleocapsids themselves are resistant to apoptosis. To confirm this hypothesis, cesium chloride (CsCl) density gradient centrifugation analysis (22) of the 4-h post-CTL cleared culture supernatant was done in order to prove that the HBcAg and HBV DNA signals described above were derived from nucleocapsid particles (Fig. 2B). Specifically, 3.88 g of CsCl was dissolved in 10 ml of cleared culture supernatant and centrifuged at 40,000 rpm in an SW41 rotor (Beckman Instruments, Inc., Palo Alto, Calif.) for 63 h at 4°C. Twenty-five 0.25-ml fractions were taken from the meniscus by pipetting. Each sample was assayed for HBV DNA, HBcAg, and HBeAg, and its density was obtained by measurement of the refractive index. As shown in Fig. 2B, both HBcAg and HBeAg and HBV DNA were detected in the same fractions in the gradient at a density of 1.35 g/ml, the density of native HBV core particles (20). Collectively, these results prove that the HBV DNA replicative intermediates that are resistant to CTL-induced apoptosis are located inside viral nucleocapsid particles.

The entire population of intracellular HBV nucleocapsids remains intact during hepatocellular apoptosis.

Next we investigated if all of the intracellular HBV nucleocapsids were resistant to apoptosis, including the newly formed immature capsids that contain single-stranded (SS) HBV DNA and the mature capsids that contain double-stranded (DS) HBV DNA. Total DNA was extracted from adherent living cells, nonadherent cells and cellular debris present in the supernatant, and cell-free supernatant, at different times after exposure to CTLs (Fig. 3). One-third of the total hepatic DNA extracted from the living hepatocytes, apoptotic debris, and culture supernatant from each dish was digested with HindIII and probed for HBV DNA by Southern blotting as described (13). As expected, in the absence of CTLs (Fig. 3A), virtually all of the replicative intermediates were found in the living hepatocytes for the duration of the study. In addition, there was no increase in ALT activity in the supernatants during the 24-h time period of the experiment, indicating that there was little or no spontaneous cell death in these cultures (Fig. 3A). In contrast, HBV DNA-replicative intermediates rapidly appeared in the apoptotic cellular debris, as early as 1 h after the addition of CTLs (Fig. 3B), concomitant with an increase in ALT activity and the level of DNA-replicative intermediates in the supernatant. Importantly, both the SS and DS forms were present in the debris and supernatants at ratios similar to the ones observed in the living cells in the corresponding control cultures to which no CTLs had been added (compare lanes S in Fig. 3B with lanes L in Fig. 3A from the corresponding time points), indicating that the immature and mature capsids were equally resistant to apoptosis. By 4 h, the ALT activity was maximal, and the DNA-replicative intermediates were highly detectable in the cell-free supernatants, reflecting the lysis of apoptotic hepatocytes. Virtually all of the hepatocytes were dead and had detached by this time because the integrated transgene and replicative intermediates were almost undetectable in the living cell fraction. In contrast, both the transgene and replicative forms were still present in the apoptotic debris at this time and at the 8-h sampling interval. By 24 h, however, the integrated transgene was eliminated from the apoptotic debris, while all of the replicative forms persisted in the debris as well as in the cell supernatant, indicating that apoptosis was complete, that the transgene had been destroyed, and that the mature and immature capsids were resistant to this process. The released capsids were stable for at least 5 days after addition of the CTLs (Fig. 3C). The dominance of mature DS DNA intermediates in the apoptotic debris at that late time reflects the dominance of the same forms in the living cells, presumably reflecting a spontaneous decrease in the rate of immature-capsid assembly relative to the rate of capsid maturation in the cells after several days in culture.

Collectively, these results indicate that HBV nucleocapsids and their content of DNA-replicative intermediates are exceptionally resistant to the proteolytic and endonucleolytic events that the CTLs trigger to destroy the host cell. This is consistent with the previously documented stability of empty HBV nucleocapsid particles to digestion by proteolytic enzymes in vitro (24), and it extends those observations to the universe of caspases, endonucleases, and other undefined destructive activities that are operative within apoptotic cells. Accordingly, the data indicate that the destructive effector function of the CTL response per se is unable to purge the replicating HBV genome from infected tissue. Since HBV is efficiently controlled, however, by the immune response during natural infection (7), other immune effector functions that control this infection must be activated in vivo. For example, it has been demonstrated that HBV replication in the livers of these transgenic mice is very efficiently controlled by inflammatory cytokines produced by antigen-specific CTLs (10, 12, 27), CD4⁺ T cells (8), and other non-antigen-specific inflammatory events (5, 6, 11) under minimally cytopathic conditions. Thus, it is likely that similar noncytolytic antiviral mechanisms are principally responsible for viral clearance during HBV infection. The current results also emphasize the relative inefficiency of the cytodestructive effector function of the CTL response by illustrating that even when CTLs kill infected cells, additional downstream events, perhaps antibody opsonization and phagocytosis, are probably needed to eliminate the replicating viral genomes that are released when the cells are destroyed.