An autophagosome-based therapeutic vaccine for HBV infection: a preclinical evaluation

All experimental protocols were approved by the Institutional Animal Care and Use Committee of Southeast University.

C57BL/6 female mice were purchased from the Comparative Medicine Center, Yangzhou University. All mice were bred and maintained in specific pathogen-free conditions. HepG2.2.15 and HepG2 cell lines were gifts from Dr. Jianqiong Zhang (Medical School of Southeast University, China). All the cells were cultured in complete medium made of DMEM or RPMI 1640 (Gibco, USA) supplemented with 10% heat-inactivated FCS (Hyclone, USA), 100 U/ml penicillin, 0.1 mg/ml streptomycin (Beyotime Institute of Biotechnology, China).

Naïve C57BL/6 mice (female, 6–8 weeks old) were injected with 10 μg of pAAV/HBV1.2 plasmid DNA (containing the HBV full-length genomic DNA) via the hydrodynamic injection as described [18].

HepG2.2.15 cells containing transfected HBV full-length DNA [17] or HepG2 cells were treated with 200 nmol/L Bortizomib (Millennium pharmaceuticals, USA) alone or in combination with 100 nmol/L Rapamycin (Enzo Life Sciences, China) or 30 mmol/L NH₄Cl for 18 h, DRibbles containing autophagosomes were prepared from the culture media as described [13],[14]. Morphology analysis of HBV⁺ DRibbles was done under transmission electron microscopy. HBsAg in HepG2.2.15 cell lysates was detected by western blot analysis using anti-HBsAg antibody (Santa Cruz, Biotechnology, Inc., USA). LC3 in both HepG2.2.15 cell lysates and HBV⁺ DRibbles was determined by western blot analysis using the polyclonal LC3B antibody (Cell Signal Technology, USA,1:1000), the HRP-labeled goat anti-rabbit IgG secondary antibody (1:5000) and a chemiluminescence kit (Multisciences Biotech Co., Ltd., China). Levels of HBsAg and HBeAg in HBV⁺ DRibbles were measured by ELISA (Shanghai Kehua Bioengineering Co., Ltd., China).

C57BL/6 mice were immunized with different doses of HBV⁺ DRibbles or HBV⁻ DRibbles (100, 30, 10 μg total protein per mouse) or PBS via intranodal injection. Seven days later, 2 × 10⁵ lymphocytes per well were harvested and re-stimulated with HBV antigens or peptides (HBc129-140: PPAYRPPNAPIL; HBs190-197: VWLSVIWM) (ChinaPeptides Co., Ltd., China) for 24 h. The number of IFN-γ producing cells was detected by ELISPOT assay (Laizee Biotech Co., Ltd., China).

To determine whether HBV⁺ DRibbles elicited cell response and killed HBV infected hepatocytes, lymphocytes were harvested from HBV⁺ DRibbles vaccinated mice at day 7 and used as effector T cells; hepatocytes were collected from mice injected with pAAV/HBV1.2 plasmid DNA as the HBV-expressing target cells. Lymphocytes (2 × 10⁶/well) were co-incubated with target cells (2 × 10⁴/well). The supernatants were collected for detection of AST by clinical chemistry analyzer after 24 h or for detection of IFN-γ by ELISA (eBioscience, USA) after 72 h.

C57BL/6 mice were immunized with HBV⁺ DRibbles, HBV⁻ DRibbles or PBS via intranodal injection as described above. Seven days later, 10 μg of pAAV/HBV1.2 plasmid DNA was injected via tail vein [18]. Serum samples were collected at day 14 for detection of HBeAg by ELISA and HBV genomic DNA with real-time PCR. Liver tissues were collected and embedded in paraffin. Intracellular HBcAg expression was detected by immunohistochemical staining with the rabbit anti-HBcAg antibody (1:100) using two-step method (Zhongshan Goldenbridge Biotechnology Co., Ltd., China). Percentages of HBcAg⁺ hepatocytes were calculated by counting at least 500 cells in five vision fields by two investigators for each sample.

Mice were vaccinated with 30 μg HBV⁺ DRibbles or PBS as above mentioned. On day 7, after injection of pAAV/HBV1.2 DNA into the mice, 200 μg anti-CD4 mAb (clone GK1.5, rat IgG2b) or/and 20 μg anti-CD8 mAb (clone 2.43, rat IgG2b) were administered intraperitoneally, and the depleted condition was maintained by repeated injections of the monoclonal antibody every 3 days for total of 3 injections. At day 14, HBeAg level and HBV DNA copy number in the serum samples and intrahepatic HBcAg expression in liver tissues were detected as described above.

C57BL/6 mice were injected hydrodynamically with pAAV/HBV1.2 plasmid DNA and were vaccinated with 30 μg HBV⁺ DRibbles, 2 μg HBsAg with Al(OH)₃ adjuvant (Center for Disease Control of Jiangsu Province, China) or PBS via intranodal injection 2 days later. Then, two intramuscular boost injections (2 μg of HBsAg with Al(OH)₃ adjuvant or 30 μg of HBV⁺ DRibbles with αAl₂O₃ nanoparticles adjuvant [19]) were performed on day 5 and 7. The lymphocytes were isolated 7 days after first vaccination and re-stimulated with recombinant HBV antigens and peptides as described. The number of IFN-γ producing cells in 2 × 10⁵ lymphocytes was detected by ELISPOT. The serum samples were collected on day 7 and 14 after first vaccination for measurement of ALT, AST, HBeAg , HBsAg, HBV DNA. The serum anti-HBsAg antibodies were detected by ELISA on day 14 after first vaccination. Liver tissues were collected on day 14 after first immunization for detection of percentage of HBcAg⁺ hepatocytes as mentioned above. The liver sections were stained with hematoxylin-eosin.

C57BL/6 mice were injected hydrodynamically with 10 μg pAAV/HBV1.2 plasmid DNA to establish HBV tolerant mice [18]. Sixty days later, tolerance toward HBsAg was noted in HBV carrier mice. The HBsAg tolerant mice were selected and then immunized with above HBV⁺ DRibbles or HBsAg or PBS on day 61, respectively. Two intramuscular boost injections with the same dose of vaccines as above were done on day 63 and 65. Lymphocytes were harvested from the immunized mice and re-stimulated with HBV antigens and peptides on day 67 as described. The number of IFN-γ producing cells in 2 × 10⁵ lymphocytes was measured by ELISPOT. In addition, at day 68, a total of 1 × 10⁸ lymphocytes from vaccinated HBsAg tolerant mice were adoptive transferred intravenously into each of the new set of 60-day HBV carrier mice. Serum samples were collected on day 3, 7, 14 and 21 after adoptive transfer for detection of ALT, AST, HBsAg and HBV DNA. The serum anti-HBsAg antibodies were measured by ELISA on day 21 after adoptive transfer. Liver tissues were collected on day 21 for detection of percentage of HBcAg⁺ hepatocytes. The liver sections were stained with hematoxylin-eosin.

One-way ANOVA or two-way ANOVN with Bonferronic post test and the two-tails paired t test were performed using GraphPad Prism5.0, (GraphPad Software Inc., San Diego, CA). A P value of <0.05 was considered statistically significant.

First, we investigated whether DRibbles contained HBV antigens when they were prepared from HepG2.2.15 cells that were transfected with HBV DNA and stably expressed HBV antigens. HepG2.2.15 cells were treated with Bortizomib, Rapamycin or NH₄Cl at the pre-determined optimal concentrations; whole cell lysates and DRibbles were prepared as described and subjected to western blot analysis using antibodies against HBsAg and LC3 [13]. The major HBsAg protein did not differ after different treatments (Figure 1A). Interestingly, Bortizomib treatment led to appearance of HBsAg of lower molecular weight than full-length products and addition of NH₄Cl further increased the appearance of these bands (Figure 1A). Because these bands appear only after inhibition of proteasome function, they are most likely representatives of DRiPs of HBsAg. No significant conversion of LC3 protein was found in HepG2.2.15 cells treated with Bortizomib, Rapamycin, or combination of both. However, accumulation of LC3-II in cell lysates and predominant LC3-II in the DRibbles was found when NH₄Cl was included in addition to Bortizomib and Rapamycin treatment (Figure 1B,C) [20],[21]. Transmission electron microscopy analysis showed that DRibbles had a double-membrane structure and their size was ranged from 100 to 1000 nm (Figure 1D). ELISA analysis showed that the amount of HBsAg and HBeAg in the DRibbles were doubled when all three drugs were combined (Figure 1E,F). These data indicated that combination of Bortizomib, Rapamycin, and NH₄Cl could induce efficient sequestration of both HBsAg and HBeAg into autophagosomes.

To determine the optimal dose of HBV⁺ DRibbles to elicit HBV antigen-specific immune responses, different doses of HBV⁺ DRibbles or HBV⁻ DRibbles were administrated into the lymph nodes of naïve C57BL/6 mice (Figure 2A). We chose the intranodal injection of DRibbles as the route of antigen delivery because our preliminary experiments indicated direct injection of DRibbles into lymph nodes resulted the greatest DCs recruitment and T-cell activation. ELISPOT assay revealed that spleen and lymph nodes from mice injected with 30 μg of HBV⁺ DRibbles contained the highest number of HBsAg or HBcAg specific IFN-γ producing cells (Figure 2B). Injection of 10 μg of DRibbles was insufficient, while 100 μg showed a high-dose suppression. To further determine whether immunogenicity and xenoreaction of DRibbles derived from human HepG2 cells (HBV⁻ DRibbles) lead to immune reactive to HBV antigens, C57BL/6 mice were also immunized with titrated amount of HBV⁻ DRibbles. Noticeably, lymphocytes from 30 μg HBV⁺ DRibbles vaccinated mice generated much higher number of IFN-γ producing cells than that from 30 μg HBV⁻ DRibbles vaccinated mice when re-stimulated with either HBV antigens or HBV peptides (Figure 2C). Thus, HBV⁺ DRibbles effectively induced both HBsAg- and HBcAg-specific and HBc_129–140, HBs_190–197-specific cell-mediated immune responses. Thirty μg total proteins seem to be optimal dose for the HBV⁺ DRibbles vaccine administrated via intranodal injection in mice.

To evaluate whether 30 μg of HBV⁺ DRibbles could protect immunized mice from establishment of HBV replication via HBV plasmid DNA injection, C57BL/6 mice were vaccinated with PBS or different doses of HBV⁺ DRibbles or HBV⁻ DRibbles and were challenged with hydrodynamic injection of HBV DNA 7 days later (Figure 2A). Serum HBV DNA and HBeAg levels were used as the indicators of viral replication and production. Mice immunized with 30 μg HBV⁺ DRibbles, but not naïve mice or mice immunized with HBV⁻ DRibbles and other doses of HBV⁺ DRibbles, were protected from HBV DNA challenge (Figure 2D,E). To further corroborate this data, liver tissues were harvested and intrahepatic expressing of HBcAg was quantified by immunohistochemical staining. Similar to the results of HBeAg and HBV DNA, the percentage of HBcAg⁺ hepatocytes in the mice immunized with 30 μg HBV⁺ DRibbles was significantly lower than control mice injected with HBV⁻ DRibbles or PBS (Figure 2F). Moreover, there were no significant difference between the mice immunized with HBV⁻ DRibbles and those with PBS. Our results showed that HBV⁺ DRibbles vaccination, but not HBV⁻ DRibbles, could effectively reduce HBV replication and expression.

To investigate the requirement of CD4⁺ helper T cells or CD8⁺ effector T cells for the control of HBV replication, anti-CD4 mAb or/and anti-CD8 mAb were administrated intraperitoneally on the same day as HBV DNA injection into HBV⁺ DRibbles-vaccinated mice. The monoclonal antibodies were injected repeatedly every 3 days to maintain the depleted condition (Figure 3A). Results showed again that non-depleted mice immunized with HBV⁺ DRibbles had a reduced level of serum HBeAg (Figure 3B), DNA copy number (Figure 3C) and HBcAg⁺ expressing hepatocytes (Figure 3D). The DNA copy number in mice depleted of CD8⁺ effector T cells were not significantly different from that of control non-immunized naïve mice. Depletion of both CD4⁺ and CD8⁺ effector T cells did not further increase the serum HBeAg or DNA copy number beyond that of mice received anti-CD8 antibody (Figure 3B,C). These data indicated the critical requirement for priming of anti-HBV CD8⁺ T cells for viral clearance.

To examine the effect function of T cells from HBV⁺ DRibbles vaccinated mice, lymphocytes from HBV⁺ DRibbles immunized mice were co-cultured with hepatocytes isolated from HBV DNA injected mice (Figure 3A). Results showed that significantly higher levels of IFN-γ were detected in the culture supernatants when effector T cells were incubated with HBV-expressing hepatocytes than control hepatocytes (Figure 3E). Moreover, AST detection showed that significantly higher levels of AST enzyme release were found in HBV-expressing hepatocytes than control hepatocytes (Figure 3F). These data suggested that HBV⁺ DRibbles elicited strong cell immune response that was capable to produce IFN-γ and kill HBV-expressing hepatocytes isolated from mice in vitro.

To test whether HBV⁺ DRibbles could be used as therapeutic vaccine in a setting that mimicking acute HBV infection and to compare the efficacy of HBV⁺ DRibbles with recombinant protein HBsAg vaccine, C57BL/6 mice were first injected with pAAV/HBV1.2 DNA to establish HBV acute infection and then were immunized with HBV⁺ DRibbles or HBsAg or PBS 2 days later. Two times boost immunizations were administrated by intramuscular injection every 2 days. Here, α-Al₂O₃ nanoparticles adjuvant was used in our HBV⁺ DRibbles vaccine due to its ability to greatly increase the efficiency of cross-presentation and anti-tumor response of cancer vaccines [19] (Figure 4A).

Consistent with previous results, lymphocytes harvested from HBV⁺ DRibbles-immunized mice generated a significantly higher number of IFN-γ producing cells than that from PBS or HBsAg immunized mice when re-stimulation ex vivo with HBV antigens or peptides (Figure 4B,C,D). The serum levels of HBeAg in HBV⁺ DRibbles-immunized mice were greatly reduced as compared to both HBsAg vaccinated mice and control mice, and no statistical difference was found between HBsAg vaccinated mice and the control mice (Figure 4E). We found that the levels of serum HBsAg in HBV⁺ DRibbles immunized mice were significantly lower than that of control mice without vaccination, but much higher than that of HBsAg immunized mice (Figure 4F). Detection of anti-HBsAg antibody in serum from immunized mice showed that HBsAg vaccination induced significantly higher levels of anti-HBs compared with HBV⁺ DRibbles vaccination (Additional file 1: Figure S1A). Serum anti-HBsAg antibody was detected only in two of six mice received HBV⁺ DRibbles immunization. Most importantly, HBV DNA copy number in the sera (Figure 4G) and percentage of HBcAg⁺ hepatocytes in the liver tissues (Figure 4H) in HBV⁺ DRibbles-immunized mice were greatly reduced as compared to control mice received either no vaccine or HBsAg vaccine. In addition, sera ALT and AST levels were determined and appeared not to be influenced by the vaccinations (Additional file 1: Figure S1B,C) and the livers from HBV⁺ DRibbles vaccinated mice showed normal architecture and a mild inflammatory responses (Additional file 1: Figure S1D). Thus far, our studies indicated that HBV⁺ DRibbles were more effectively than HBsAg, at eliciting therapeutic cell-mediated immune responses in an acute HBV infection mouse model established by hydrodynamic injection of HBV plasmid DNA and did not cause obvious liver damage.

Next, we investigated whether HBV⁺ DRibbles immunization could induce therapeutic immunity and break tolerance in a HBV tolerant mouse model. HBsAg tolerant mice were established by hydrodynamic injection of HBV DNA as described [18] (Additional file 2: Figure S2A). Two months after DNA injection, mice that are HBsAg positive were inoculated with HBV⁺ DRibbles or HBsAg or PBS. Seven days after first immunization, lymphocytes were harvested and re-stimulated by HBV antigens and peptides (Figure 5A). The data showed that HBV⁺ DRibbles, but not HBsAg, induced HBV-specific IFN-γ spot-forming cells when re-stimulated ex vivo with recombinant HBsAg or HBcAg proteins or MHC I binding peptides derived from these antigens (Figure 5B,C). Since the level of antibody response in DRibbles-vaccinated mice in the acute ‘HBV infected’ mouse model was very low, we decided to examine whether immune cells rather than serum conferred the control of HBV replication using an adoptive transfer model. To that end, immunized lymphocytes harvested from spleens and lymph nodes of immunized mice were adoptively transferred to a new set of HBsAg tolerant mice that received HBV DNA injection 60 days before. The levels of serum HBsAg, HBV DNA (Figure 5D,E) were markedly decreased and HBcAg⁺ hepatic cells were greatly reduced in HBsAg tolerant mice after adoptive transfer of lymphocytes from HBV⁺ DRibbles vaccinated mice (Figure 5F). In contrast, the mice received lymphocytes from HBsAg vaccinated mice did not exhibit any measurable effect. Interestingly, anti-HBsAg antibodies were detected in mice received lymphocytes from DRibbles, but not PBS and HBsAg immunized mice (Additional file 2: Figure S2B). Furthermore, the serum ALT and AST levels in the mice received lymphocytes from HBV⁺ DRibbles immunized mice were similar to that of received lymphocytes from non-immunized control mice (Additional file 3: Figure S3A,B) and the livers of the mice receiving HBV⁺ DRibbles-treated lymphocytes showed normal architecture with a mild inflammatory infiltrates (Additional file 3: Figure S3C). The results indicated that HBV⁺ DRibbles were able to reverse the tolerance to HBsAg and induce therapeutic T cells, and these HBV-specific T cells adoptively transferred into HBsAg tolerant mice did not cause severe liver damage.