|Home | About | Journals | Submit | Contact Us | Français|
The prevalence of hepatitis B virus (HBV) infection in Asia and sub-Sahara Africa is alarming. With quarter of a billion people chronically infected worldwide and at risk of developing liver cancer, the need for a prophylactic or therapeutic vaccination approach that can effectively induce protective responses against the different genotypes of HBV is more important than ever. Such a strategy will require both the induction of a strong antigen-specific immune response and the subsequent deployment of immune response towards the liver. Here, we assessed the ability of a synthetic DNA vaccine encoding a recombinant consensus plasmid from genotype A through E of the HBcAg, to drive immunity in the liver. Intramuscular vaccination induced both strong antigen-specific T cell and high titer antibody responses systematically and in the liver. Furthermore, immunized mice showed strong cytotoxic responses that eliminate adoptively transferred HBV-coated target cells. Importantly, vaccine-induced immune responses provided protection from HBcAg plasmid-base liver transfection in a hydrodynamic liver transfection model. These data provide important insight into the generation of peripheral immune responses that are recruited to the liver- an approach that can be beneficial in the search for vaccines or immune-therapies to liver disease.
Hepatocellular carcinoma or liver cancer is the third most common cancer and the most deadly, killing most patients within 5 years of diagnosis. About 600,000 new cases arise each year. One of the major causes and risk factors for liver cancer is infection by hepatitis B1, 2
Hepatitis B virus (HBV) is a hepadnavirus that infects human hepatocytes and causes acute and chronic hepatitis B infection. Human HBV has been divided into four major serotypes (adr, adw, ayr and ayw) according to the antigenic determinants presented on the surface protein, and subdivided into eight genotypes (A through H)3.
Although a vaccine against HBV has existed for over 3 decades, HBV remains a major epidemic, especially among people of Asian and African descent. One third of the world’s population has been infected with HBV, with 370 million people chronically infected with the virus and at risk of developing liver cirrhosis or hepatocellular carcinoma4, 5. Currently, the only therapies available for chronically infected individuals are interferon-α and nucleoside analogue treatments, which function by controlling viral replication but unfortunately do not clear infection. Interferon-α can prevent viral replication in only 30% of patients and does so with undesirable side effects. On the other hand, nucleoside analogues are much more effective at inhibiting viral replication but prolonged treatment often results in the emergence of escape mutants6. The failure of protection by current HBV vaccines in 15% of individuals and the difference in response to treatment in chronically infected people may be due to the genetic divergence among different HBV genotypes7–9, as well as the inability to drive strong HBV-specific cytotoxic lymphocytes (CTLs) to the liver.
In the hopes of addressing these inadequacies with current approaches, we constructed a plasmid based on the HBV core (HBcAg)-specific consensus sequence from the Asian and African genotypes of HBV with genetic modifications that improve expression of the inserts. This includes codon and RNA optimization, as well as additional modifications to elicit maximum in vivo expression10. HBV core protein was chosen as it is the major viral determinant of HBV persistence, with one of the major differences between acute and chronic human HBV infection being the detection of CTL against HBcAg in circulation11–13. Previous data from several groups suggest that these CTLs play a crucial role in the clearance of HBV during acute infection. In addition, recent findings suggest that the induction of both cellular and humoral immunogenicity against HBV antigens are important for controlling chronic infection14. Studies from our laboratory and other independent groups have shown that plasmid DNA delivery by electroporation (EP) represents an important strategy for enhancing cell or antibody mediated immune responses15–18, which we examined in this study. We report that a synthetic construct is able to induce a strong antigen-specific T cell response that mediates cytotoxicity systematically and induces robust T cell migration into the liver. In addition, a high titer antibody response capable of recognizing a native HBcAg was observed. Importantly, immunized mice were able to clear transfected HBV antigen–expressing hepatocytes in vivo illustrating the functional nature of these responses in clearing antigen from the liver.
A HBV genotype A, B, C, D and E Core consensus nucleotide sequence was constructed by generating consensus sequences of core genes for each genotype and then generating a consensus sequence of all five genotype consensuses, thus avoid biasing toward heavily sequenced genotypes. The sequences were aligned using clustal X software to develop the final HBcAg consensus sequence. Once the group consensus sequence was obtained, we introduced several modifications, including the addition of a highly efficient IgE leader sequence and a C-terminal HA tag, and the construct was RNA and codon optimized. The synthesized HBcAg was digested with EcoRI and NotI, and cloned into the expression vector pVAX (Invitrogen) under the control of the cytomegalovirus immediate-early promoter and this construct was then named as pMCore.
Expression of pMCore was detected using TNT® Quick Coupled Expression of Transcription/Translation System containing 35S-methionine (Promega, Madison, WI). The synthesized gene product was immuno-precipitated using an anti-HA monoclonal antibody targeting an encoded HA epitope, Clone HA-7 (Sigma-Aldrich). The immuno-precipitated protein was electrophoresed on a 12% SDS-PAGE gel and subsequently fixed and dried. The synthesized protein with incorporation of radioactive 35S was detected by autoradiography.
Rhabdomyosarcoma (RD) cell lines were transfected with pMCore using TurboFect™ (Thermo Scientific) according to manufacturer’s guidelines. The cells were first fixed with 2% formaldehyde and then assayed for protein expression. The fixed cells were incubated with rabbit monoclonal HA tag (Invitrogen) diluted in ‘primary standard solution’ (0.1% BSA, 0.2% saponin, 0.02% sodium azide) for an hours at room temperature. The cells were subsequently incubated with DyLight 594-labeled anti-rabbit secondary antibody (Thermo Scientific) for 20 minuets at room temperature. Images were obtained using a Zeiss Axiovert 100 inverted confocal microscope. Analysis and quantification of florescence intensities were conducted using Image J software (NIH, Rockville, MD).
Six to eight week old female Balb/c mice were purchased from Jackson Laboratories. Animals were maintained in accordance with the National Institutes of Health and the University of Pennsylvania Institutional Care and Use Committee (IACUC) guidelines. For DNA immunization studies, eight animals were divided in to two groups. Each animal in the immunized group received a total of three intramuscular immunizations of 30 μg of pMCore two weeks apart. Each immunization was accompanied by in vivo electroporation with the CELLECTRA® adaptive constant current electroporation device (Inovio Pharmaceuticals, Blue Bell, PA). Two 0.2 Amp constant current square-wave pulses were delivered through a triangular 3-electrode array consisting of 26-gauge solid stainless steel electrodes completely inserted into muscle. Each pulse was 52 milliseconds in length with a 1 second delay between pulses.
Splenocytes were isolated as described elsewhere19. In brief, mice were sacrificed one week after the last immunization and spleens were harvested, placed in R10 media (RPMI media supplemented with 10% FBS and 1x Anti-Anti). The spleens were individually crushed, strained with a 40μM cell strainer and treated with 1mL ACK lysis buffer for 5 min to lysis erythrocytes. The splenocytes were resuspended in a complete R10 media and used for further immunological assays.
Lymphocytes from liver were obtained as described elsewhere 20. Briefly, each liver was perfused by directly injecting 1mL of PBS into the hepatic vein of each mouse. Livers were harvested, crushed and resuspended in 5mL of 44% isotonic percoll. The mixtures were underlied with 3mL 66% isotonic percoll and centrifuged for 20 minutes at 2000rpm for gradient separation. Lymphocytes were collected and washed in 10 mL R10 and treated with ACK lysis buffer as necessary.
IFN-γ ELISpot was performed as previously described21. Splenocytes were stimulated with two pools of 15-mer peptides spanning the entire length of pMCore and over lapping by 8 amino acids. 200,000 splenocytes in R10 media were plated in a 96 well IFN-γ capture antibody (R&D system) coated plate and stimulated overnight in the presence of a specific peptide pool at 37°C in 5% CO2. Cells were washed out and plates were incubated overnight with biotinylated anti-mouse IFN-γdetection antibody (R&D system). Streptavidin-alkaline phosphatase and 5-bromo-4-chloro-3′-indolylphosphate p-toluidine salt and nitro blue tetrazolium chloride were subsequently used to develop spots. Spots were counted using an automated ELISPOT reader (CTL Limited).
A subset of splenocytes was resuspended in R10 media at a concentration of 107 per mL and 100μL was plated onto a 96 well round bottom plate. 100μL of media containing pMCore pooled peptides or 10 ng/ml PMA (Sigma, St. Louis, MO, USA) and 500ng/ml ionomycin (Calbiochem, Novabiochem, La Jolla, CA, USA) mix as a positive control or 0.1% dimethyl sulfoxide (Sigma, St. Louis, MO, USA) as a negative control. All wells contained 5uL/mL of two protein transport inhibitors, brefeldin A (GolgiPlug) and monensin (Golgistop) (All from BD Bioscience). The cells were incubated at 37°C in 5% CO2 for 5 hours and stained with LIVE/DEAD® Fixable Dead Cell Stain (invitrogen) for 10min at 37°C. Extracellular staining was performed using antibodies specific to mouse CD3, CD4 and CD8. Spleenocytes were then permeabilized and washed using BD Cytofix/Cytoperm™ and Perm/Wash™ (BD Bioscience) respectively. The cells were then stained intracellularly with antibodies to mouse Interferon-gamma and Tumor Necrosis Factor- alpha.
Conjugated anti-mouse antibodies were used during the extracellular and intracellular staining including: CD3- Phycoerythrin/Cy5 (PE/Cy5), CD4- peridinin chlorophyll protein (PerCP), CD8- allophycocyanin (APC), IFN-γ-Alexa Fluor 700, TNF-a- fluorescein isothiocyanate (FITC) and IL-2-phycoerythryin cyanine (PE) (all from BD Biosciences, San Jose, CA).
The T cell proliferation assay using CFSE labeling has been previously described22. Briefly, isolated splenocytes were stained with the carboxyfluorescein diacetate, succinimidyl ester (CFDA-SE) Cell Tracer Kit (Invitrogen) as per the manufacturer’s instructions. Stained cells were washed three times with saline and plated in a 96-well U-bottomed plate in a total volume of 200 μL of stimulating media. The cells were incubated at 37°C for 96 hours. After 48 hours, 50% of the culture media were removed and replaced with fresh R10.
An in vivo cytotoxicity assay was performed as previously described23,24. Briefly, splenocytes from naïve mice were stained with either 1μM or 1nM CFDA SE (invitrogen). The labeled splenocytes were then coated with indicated peptides (1μM) and 107 cells of each population intravenously injected into naïve or immunized mice. After 24 or 90 hours cells from the spleen and liver were isolated and analyzed by flow cytometry. The percent killing was calculated as follows: 100 − ([(% relevant peptide pulsed in infected/% irrelevant peptide pulsed in infected)/(% peptide pulsed in uninfected/% irrelevant peptide pulsed in uninfected)] × 100).
High-binding ELISA plates (Costar, Corning, NY) were coated with 1}μg/ml HBcAg protein in PBS, at 4°C for 24 hours and then were washed with 0.1% PBS-Tween then and blocked with PBS containing 1% BSA for 2 hours at room temperature. Serially diluted serum samples were added to the wells and incubated for 1 hour at room temperature. After washing, bound serum Antibody was revealed by HRP-labeled goat anti-mouse IgA or IgG. The peroxidase-conjugated antibodies were detected using tetramethylbenzidine (Sigma-Aldrich) as the substrate, and OD values at 450 nm were measured with the Multiscan ELISA Plate Reader.
The in vivo clearance assay was performed as described elsewhere 25,26. Briefly, immunized mice were injected intravenously with 100μg of plasmid in 2mL (about 10% volume of the mouse weight) of Ringers solution within a period of 7 seconds to transiently transfect the liver. The expression or clearance of the plasmid was determined by staining the liver with anti-HA monoclonal antibodies.
Serum alanine aminotransferase (ALT) activity was measured using an absorbance-based assay (Stanbio Laboratory) on a BioTek Synergy 2 microplate reader. Results are reported as units per liter (U/L) and represent the amount of enzyme that oxidizes one μmol/L of NADH per minute.
A consensus sequence of HBcAg was generated from 5 different genotype (A through E) gene sequences. The sequences were collected from different countries to avoid sampling bias towards heavily sequenced genotypes. As shown in Fig 1A, there was an observed relative closeness of the multi-genotype consensus HBcAg sequence to all sampled sequences from different genotypes. After the consensus sequence was generated, several modifications were performed to increase the antigen expression levels from plasmid as described by our laboratory for other plasmid-based vaccines10, 19 (Fig 1B).
The HBcAg protein was expressed by transfected pMCore DNA plasmid containing the core gene of hepatitis B (Fig 1B). In vitro translation assay on lysate showed detectable HBcAg at an expected molecular weight of 28 kDa ( Figure 2A). The expression was further confirmed using anti-HA tagged monoclonal antibody by immunofluorescent assay. We took advantage of confocal imaging to visualize HBcAg in the cytoplasm and around the nucleus of tranfected RD cells as shown in Figure 2B. The expression pattern confirmed that a DNA plasmid carrying the core gene could be expressed at highly expressed in different cells in vitro.
To better evaluate the generation of T and B cell immune responses, we immunized Balb/c mice and measured both responses in various peripheral tissues. Mice received three intramuscular immunizations of 30μg of pMcore or pVax followed by electroporation as depicted in the immunization scheme (Fig. 3A). One week after the final immunization, pMCore immunized mice showed evidence of strong HBcAg T cell responses as identified by IFN-γ ELISPOT assay following ex vivo stimulation. Figure 3B clearly shows the dominant epitopes are biased towards peptide pool 2. The average HBcAg-specific IFN-γ T cell response induced were robust at 2000 (± 210) SFU per million splenocytes. Interestingly, intracellular staining of stimulated splenocytes revealed that both CD4+ and CD8+ cells produce almost the same amount of antigen-specific IFN-γ and TNF-α with about 0.4% of the T cells being double positive for both cytokines (Figure 3C and 3D). The comparable cytokine production between both T cells did not predict their ultimate proliferation capacity. After 4 days of stimulation with antigen specific peptides, the CD8+ T cells proliferated more than 2 fold higher than the CD4+ cells (Figure 3E) showing a clear CD8 T cell bias in the response.
To further explore the immune response induced in pMCore-immunized mice, we analyzed antigen-specific IgG and IgA responses by B cell ELISpot as well as in ELISA using splenocytes and sera, respectively, collected following vaccination. A high IgG and IgA titer was observed in the sera of immunized mice when compared to control animals. B cell ELISpot from immunized mice showed HBcAg-specific IgG and IgA at approximately 200 SFU and 100 SFU per million cells respectively. Figure 3F, illustrates activation of the B cell compartment by immunization. Our synthetic HBcAg plasmid effectively induced antigen-specific cellular and humoral responses after 3 immunizations.
It is hypothesized that increasing functional anti-viral T cell effectors in the liver will be important to clear chronic HBV infection. However, few studies on HBV vaccines have attempted to address this issue. Here, we examined the cytokine producing capabilities of intrahepatic antigen-specific T cells after DNA immunization. Both CD4 and CD8 T cells isolated from the liver produce IFN-γ and TNF-α when stimulated in vitro with HBcAg peptide (Fig 4A and B). While the CD4 T cells show a high percentage of double producers, the CD8 showed little to no IFN-γ+TNF-α+ producing cells. Instead majority of the CD8 T cells produced only IFN-γ or TNF-α. We also observed enrichment of HBcAg-specific CD4 T cells in the liver as compared to the spleen. The percent HBcAg-specific CD4 T cell producing IFN-γ and double positive CD4 T cells in the resting liver were 4 and 2.8 fold higher than that observed in the spleen respectively. Conversely, peripheral CD8 T cells were confirmed to be better double producers than liver resident CD8 T cells. We also observed antibody-producing capabilities of liver resident B cells from immunized mice. Interestingly, the liver as a mucosal organ produced higher antigen-specific IgA than IgG (Figure 4C), an important observation that has not been previously been studied.
Next we assessed the ability of HBV-specific CD8 T cells induced after DNA immunization to specifically eliminate target cells in vivo. As previously described, human CTLs that target the core antigen are important in acute clearance of HBV versus chronic infection. One week after the final immunization, 4 mice from each of the two groups, pVax or pMCore immunized, were adoptively transferred with target splenocytes that had either been pulsed with HBcAg (relevant) or HCV-NS3/4A (irrelevant) peptides. By gating on CFSE labeled splenocytes to track killing, we observed that the pMCore vaccinated mice were able to induce strong specific killing of antigen-pulsed target cells as shown in Figure 5. Average percent killing observed in the spleen was about 83% while the average in the liver was 76%, showing that vaccine-induced CTLs that migrate to and are retained in the liver are capable of killing HBV peptide pulsed target cells. The percent killing in the spleen was comparable to previous studies that reported HBcAg-specific CTL responses using in vitro cytotoxicity assays14. To our knowledge, this is the first study to show induction of HBcAg-specific CTL responses in the liver, by any method and specifically by systemic immunization. This data provides evidence that peripheral immunization can induce effector cells that can migrate to the liver and lyse target cells.
In the absence of a small animal model for HBV to examine immune mediated clearance, we next utilized HBcAg plasmid to transiently transfect mouse liver through direct hydrodynamic injection. This model has been described in studies to acutely transfect mouse liver with different types of viral DNA11,27, 28. Here, immunized or naive mouse livers were either transfected with pMCore or an irrelevant plasmid encoding hepatitis C antigens (HCV NS3/4A). Immunohistochemistry staining three days post transfection (Figure 6A) shows clearance of HBcAg-transfected hepatocytes as compared to the NS3/4A-transfected animal livers. CD8 T cells isolated from the pMCore hydrodynamic injected mice in Figure 6B showed a higher frequency of IFN-γ+ CD107a+, a marker of degranulation, as compared to immunized animals livers transfected with the irrelevant plasmid.
Since the clearance of pMCore-transfected hepatocytes seems to involve degranulation, it was fair to assume the killing may lead to liver damage. To examine if immunized mice were able to clear the transfected hepatocytes without inducing significant liver damage, we employed a widely used assay measuring the enzyme, alanine aminotransferase (ALT) which is an indication of liver damage when enzyme levels are observed elevated in the sera, Figure 6C. These studies showed that the specific clearance of HBcAg-transfected hepatocytes did not increase ALT levels in transfected immunized animals beyond the normal range of 5–30U/L.
Resolution of acute HBV is believed to require a strong multi-specific CD4+ T cell response to peptides encoded in the core antigen12, 29. CD8+ T cells are however, the main effector cells responsible for HBV clearance via both cytolytic and non-cytolytic pathways, induced by the production of cytokines such as IFN-γ+, TNF-α+30–32. As such, in chronically infected individuals there is a decrease in total HBV-specific CD4+ and CD8+ T cells compared to persons who successfully clear the virus. Moreover, CTL responses to the core protein appear undetectable, as patients ability to generate IFN-γ+ becomes diminished following chronic infection33, 34. Studies in animals have confirmed the elimination of both IFN-γ+ and TNF-α+ producing CD8+ cells during chronic HBV infection35. These data together with studies distinguishing HBcAg as the major viral determinant of HBV persistence 11 highlight the importance of generating a strong immune response to HBcAg to resolve HBV infection.
DNA vaccination is a malleable vaccine approach that is reestablishing itself as an important platform through improved expression and delivery technologies 36. For instance, electroporation technology delivery has not only greatly increased the magnitude of immune responses generated after DNA immunization, but also has greatly aided in reducing the overall amount of DNA required to generate such immune responses. While few studies have explored electroporation to induce immune response against surface antigens of HBV37, there have been no studies of this improved platform in the HBV model of the core antigen, which is critically important in generating immune clearance of infection.
Previous studies from other groups have shown induction of either or both cellular and humoral responses when murine38, nonhuman primates39 and humans40 when immunized with plasmids or retroviral vectors encoding the core antigen. However, the ability of these vaccines to induce and retain antigen specific immune cells in the liver and clear infected hepatocytes was not investigated. Furthermore the magnitude of the responses induced in this study appear superior to these prior approaches. In this study, we examined the proficiency of a synthetic HBcAg encoded plasmid DNA vaccine delivered in the periphery to establish antigen-specific immune response targeting to the liver resulting in clearance of HBV transfected liver cells.
The immune phenotype induced by this vaccine was interesting. Using intracellular staining we observed that antigen specific cells from the spleen of immunized mice were double positive for both IFN-γ and TNF-α. Although, the plasmid contains a core antigen, it is worthy of note the high titer HBcAg-specific IgG and IgA observed in their sera. From a therapeutic point view, it is important for the activated cells to traffic to and to then be retained in the liver. Immunized mice demonstrated increase in activated CD4 T cells in the liver as evidenced by the large percentage of HBcAg-specific IFN-γ and TNF-α double positive cells found there. Liver CD8 T cells on the other hand, were mostly single function producer cells. The importance of multiple-cytokine producing CD4 T cells in producing effector functions is well described 41 but single function CD8 and more specifically intrahepatic T cells in this regard have not been previously reported. However, single function CD8 T cells are likely effector cells. Lu et al.14, demonstrated that immunization with HBcAg DNA could induce antigen-specific cytotoxic activity in vitro. Our data using an in vivo killing assay extended their findings. The ‘in vivo killing’ assay confirmed that antigen-specific CD8 T cells in immunized mice’s spleen and liver (which were single function CD8 T cells), are able to efficiently eliminate target cells.
In order to assess the effectiveness of an HBV vaccine, a challenge model is needed. In the absence of infectious small animal model, mouse hepatocytes can be transiently transfected in vivo by plasmid DNA through a process known as hydrodynamic injection to create an “infectious-like” model. Although, this model mimics aspects of acute infection, it has been used by many independent groups to study viral kinetics27, persistence11 and clearance42. We transfected naïve and pMCore primed mice liver with the consensus DNA through hydrodynamic injection. Three days after the transfection, degranulation in spleen was evaluated by staining CD8 T cells from both groups for CD107a. The immunized group showed higher percentage of IFN-γ and CD107a double positive CD8 T cells are retained in the liver.
pMCore-Immunized mice showed complete clearance of HBcAg expressing hepatocytes but were ineffective against HCV-NS3 expressing hepatocytes, showing the level of specificity of the generated CTL. This specificity of the clearance is relevant to treatment of HBV infection, as nonspecific elimination of hepatocytes has been reported to lead to inflammation in the liver. There are concerns that cellular immune responses induced by vaccines against liver pathogens can drive hepatitis resulting in liver damage. Measuring the amount of alanine transaminaseas in the blood as means of evaluating liver injury, we observed no significant liver damage during or after the liver clearance process. This important finding suggests that antigen-positive hepatocyte clearance might be through both cytolytic and non-cytolytic process as seen during human acute HBV clearance.
In summary, a synthetic recombinant plasmid encoding HBcAg and delivered by EP can generate responses that target to liver, and such response maybe useful in a prototype HBV immune therapy. Although, further studies are needed to understand the trafficking and retention of antigen specific T cells in the liver, their ability to induce responses and clear infected hepatocytes with minimal damage by what appear to be effector T cells, as reported here, is encouraging. Further investigation of these immunization tools in HBV with relevance to liver cancer appears important.
For Disclosure purposes DBW notes that he and his laboratory have several commercial relationships with companies in the area of vaccines. These include that he receives consulting fees, or received stock ownership, for Advisory Board/ Review Board Service, or received speaking support or research support from commercial entities including: Inovio, BMS, VGXI, Pfizer, Virxsys, J & J, Merck, Sanofi Pasteur, Althea, Novo Nordisk, SSI, Aldevron, Novartis, Incyte and possibly others. No writing assistance was utilized in the production of this manuscript