Introduction
More than 257 million humans worldwide suffer from chronic hepatitis B virus (CHB) infection [
1], which can lead to liver cirrhosis and hepatocellular carcinoma (HCC), the 2
nd most common cause of cancer-related human deaths worldwide [
2]. HBV-associated human mortality increased from 0.89 million to 1.45 million between 1990 and 2013 [
3,
4] despite the availability of prophylactic HBV vaccines [
5]. Antiviral drugs reduce HBV replication, but fail to eliminate the intranuclear viral genomes from infected hepatocytes and thereby do not cure [
6‐
8].
During an acute infection CD8
+ T cells can clear HBV-infected hepatocytes through cytolysis and the release of antiviral cytokines [
9]. Persistent infections are associated with dysfunctional HBV-specific CD8
+ T cell responses [
10]. Due to sustained presence of antigens, induction of immunosuppressive cytokines [
11], regulatory T cells (Tregs) [
12], and the unique hepatic microenvironment [
13], HBV-specific CD8
+ T cells lose functions [
14,
15]. Treatment with anti-Programmed Cell Death Protein 1. (PD1) antibodies may reverse CD8
+ T cell dysfunction caused by exhaustion but has shown limited success in CHB patients [
16]. The disappointing clinical outcome of checkpoint blockade in CHB patients contrasts results obtained in a pre-clinical woodchuck model of chronic hepadnaviral infection where a combination of anti-viral drugs, checkpoint blockade and an HBV-specific DNA vaccine led to sustained immunological control of the infection or complete viral clearance [
17].
Here we describe pre-clinical results with a therapeutic vaccine to HBV, which induces potent and sustained CD8+ T cell responses that are relatively resistant to exhaustion and when combined with antiviral drugs may affect a functional cure of CHB. Implications of our findings for the development of a therapeutic HBV vaccine are discussed.
Methods
Cell lines
Human embryonic kidney (HEK) 293 cells and coxsackie adenovirus receptor (CAR)-transduced Chinese hamster ovary (CHO) cells were maintained in Dulbecco’s Modified Eagles medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and antibiotics.
Mice
Male 6 week-old C57BL/6, BALB/c and HLA-A2 transgenic (tg) (C57BL/6-Mcph1Tg(HLA−A2.1)1Enge/J) mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA). Mice were housed at the Animal Facility of the Wistar Institute and treated according to approved protocols. Unless stated otherwise experiments were conducted with groups of 5 mice, 2 or 3 times.
Production, purification, and titration of vectors
Early antigen (E)1 and partially E3-deleted AdC6 and AdC7 vectors expressing PolN, PolC or core within gD under the control of the early cytomegalovirus promoter were produced, purified and titrated as described previously [
18]. They were formulated in 2.5% Glycerol/25 mM odium chloride (NaCl)/20 mM TRIS buffer, pH 8.0.
AAV8 vectors expressing the 1.3 genome of HBV genotype D were produced in HEK 293 cells by triple plasmid (p)Helper/pAAV-1.3HBV/pAAV8
capsid transfection using vectors at ratios of 1:0.5:0.5 as described [
19]. HEK 293 cells were transfected with plasmids using polyethylenimine (PEI) at a 1:3 ratio of DNA: PEI. After a 72 h incubation in a 37 °C, 5% CO
2 incubator cells were harvested by centrifugation and washed with phosphate buffered saline (PBS). The cell pellet was re-suspended in PBS and sonicated for 5 min with 1 min on and 1 min off at 50% amplitude to release the virus. Cell pellets were treated with 50U/mL of Benzonase, 0.5% sodium deoxycholate for 30 min in a 37 °C water bath. Cell debris was removed by centrifuging and the supernatant was subjected to centrifugation in a Beckman centrifuge with a 70.1 Ti rotor at 67,000 rpm for 2 h at 18 °C with maximum acceleration and no brake over an iodixanol step gradient (from bottom to top at 60%, 40%, 25%, and 15%). AAV particles were harvested from the 40–25% interface and dialyzed against PBS in a 100 kDa Millipore Ultra-50 unit. After dialysis, purified AAV was collected, aliquoted and stored at − 80 °C.
Aliquots of the purified AAV8-1.3HBV vectors were loaded on 10% sodium dodecyl sulfate (SDS) gel (Bio-Rad, Hercules, CA) and run for 1 h using 1 × running buffer. Silver staining was performed on the gel according to manufacturer’s instruction using Pierce™ Silver Stain Kit (ThermoScientific, Rockford, IL). Gels showed the expected 3 bands of AAV viral proteins 1–3 (not shown).
SYBR green quantitative (q) polymerase chain reaction (PCR) assays were performed with the SYBR Green Master Mix from ThermoFisher (Waltham, MA) to determine the AAV titers with insert-specific primers (Forward: TGAGAGGCCTGTATTTCCCTGC; Reverse: AACCCCGCCTGTAACACGAG). Briefly, 5µL aliquot of Iodixanol purified vector was treated with DNase I in PCR buffer at 37 °C for 30 min. Treatment samples were tenfold serially diluted in duplicates and amplified using the primers for PolN or PolC, and Sybr green master mix. Standard plasmids with known genome copies were included in each plate. Viral genome (vg) copies/mL were calculated based on cycle threshold values and the plasmid-derived standard curve.
Protein expression
The Ad vectors were tested for protein expression upon transfection of HEK 293 cells or CHO cells stably transfected to express CAR. Briefly, 1 × 106 cells/flask were infected for 48 h with ~ 1000 virus particles (vp)/cell of the AdC6 or AdC7 vectors expressing gDPoLN or gDPolC. Negative control cells were transfected with a characterized vector expressing gp140 of human immunodeficiency virus; uninfected cells served as an additional negative control. Cells were collected and lysed in radioimmunoprecipitation assay buffer (RIPA) buffer supplemented with a 1% protease inhibitor (Santa Cruz Biotechnology Inc., Dallas, TX). The lysate was stored at − 80 °C until further use. A 15 µl sample of protein was resolved on 12% SDS-polyacrylamid gel electrophoresis (NuPAGE™ 4-112 Bis–Tris Gel, Invitrogen, Carlsbad, CA) and transferred to a Immobilon-FL polyvinylidene difluoride (PVDF) membrane (Merck Millipore, Burlington, MA). The membrane was blocked in 5% powder milk overnight at 4 °C. The primary antibody to gD diluted to 1:1000 in saline (clone PA1-30233, Invitrogen, Carlsbad, CA) was added for 1 h at room temperature. Membranes were washed with 1X tris-buffered saline (TBS) with Polysorbate 20 (TBST) prior to incubating with horse radish peroxidase (HRP)-conjugated goat anti-rabbit secondary immunoglobulin (Ig)G (ab6721, Abcam, Cambridge UK) for 1 h at room temperature. Membranes were washed 3 times with 1X TBS-T. The developing agent Super Signal West Pico Chemiluminescent (Thermo Fisher Scientific, Waltham, MA) was added. Membranes were shaken in the dark for 5 min, dried, and developed.
Vaccination and infection of mice
AdC6 or AdC7 vectors were diluted in sterile saline. A total volume of 200 µl containing the indicated numbers of vp was injected intramuscularly (i.m.) into the left hindlegs of mice. The AAV8-1.3HBV vector was diluted in sterile saline and injected at a volume of 300 µl into the tail vein.
Preparation of peripheral blood mononuclear cells (PBMCs)
Mice were bled from the saphenous vein and blood was collected into 4% sodium carbonate and Liebowitz’s-15 (L-15) medium. PBMCs were purified by Ficoll® Paque Plus (GE Healthcare, Chicago, IL) gradient centrifugation for 30 min at 2800 rpm. Cells were washed and seeded into 96 roundbottom well plates (0.2–1 × 106 cells per well).
Collection of lymphocytes from livers and spleens
Spleens and livers were harvested from mice. Single cell suspension was generated by mincing spleens with mesh screens in L15 medium followed by passing cells through a 70 μm Falcon™ cell strainer (Thermo Fisher Scientific). Red blood cells (RBC) were lysed by 1X RBC lysis buffer (eBioscience, San Diego, CA). To obtain hepatic lymphocytes, livers were cut into small fragments and treated with 2 mg/ml Collagenase P, 1 mg/ml DNase I (all from Roche, Basel Switzerland) and 2% FBS (Tissue Culture Biologicals, Tulare, CA) in L15 under agitation for 1 h. Liver fragments were homogenized, filtrated through 70 μm strainers and lymphocytes were purified by Percoll-gradient centrifugation and washed with DMEM supplemented with 10% FBS.
In vitro stimulation of lymphocytes
Lymphocytes were stimulated with various pools of peptides or individual peptides representing the HBV sequences present in the vaccines. Peptides were 15 amino acids in length and overlapped by 10 amino acids with the adjacent peptides. Individual peptides were diluted according to the manufacturer’s instructions in either water, DMSO, ammonia water, formic acid or N-methyl. For stimulation ~ 106 lymphocytes plated in medium containing 2% FBS and BDGolgiplug Protein Transport Inhibitor (BD Bioscience; San Jose, CA), at 1.5 μl/ml were cultured with the peptide pools or individual peptides each present at a final concentration of 2 µg/ml for 5 h at 37 °C in a 5% CO2 incubator. Control cells were cultured without peptides.
Intracellular cytokine staining (ICS) and analyses by flow cytometry
Following stimulation cells were incubated with anti-CD8- Allophycocyanin (APC) (clone 53-6.7, BioLegend, San Diego CA), anti-CD4-PerCp5 (clone Gk1.5, BioLegend), anti-CD44-Alexa Flour 700 (clone IM7, BioLegend) and Live/Dead™ fixable violet dye (Thermo Fisher Scientific) at + 4 °C for 30 min in the dark. Cells were washed once with PBS and then fixed and permeabilized with BD Cytofix/Cytoperm™ (BD Biosciences, San Jose, CA) for 20 min. Following fixation, cells were incubated with an anti-interferon (INF)-γ-FITC antibody (Clone, XMG1.2 BioLegend) at 4 °C for 30 min in the dark. Cells were washed and fixed in 1:3 dilution of BD Cytofix™ Fixation Buffer (BD Pharmingen, San Diego CA). They were analyzed by a BD fluorescent cell sorter (FACS) Celesta (BD Biosciences, San Jose, CA) and Data-Interpolating Variational Analysis (DiVa) software. Post-acquisition analyses were performed with FlowJo (TreeStar, Ashland, OR). Data shown in graph represents % of INF-γ production by CD8+ or CD44+CD8+ cells upon peptide stimulation. Background values obtained for the same cells cultured without peptide(s) were subtracted.
Tetramer staining
Lymphocytes were stained with Live/Dead™ fixable violet dye (Thermo Fisher Scientific), anti-CD8-APC (clone 53-6.7, BioLegend) anti-CD44-Alexa Flour 700 (clone IM7, BioLegend), anti-EOMES-Alexa Fluor 488 (clone Dan11mag, eBioscience), anti-PD1-BV605 (clone 29F.1A12, BioLegend), anti-LAG3-BV650 (clone C9B7W, BioLegend), anti-T-bet-BV786 (clone 4B10, BioLegend), anti-CTLA4-PE-A (clone UC10-4B9, BioLegend), anti-TIM3-Pe-Cy7-A (clone RMT3-23, BioLegend) and an APC-labeled MHC class I tetramer (NIH tetramer Facility, Emory University, Atlanta GA) corresponding to amino acids 396-404 FAVPNLQSL (peptide 55) of the HBV polymerase at + 4 °C for 30 min in the dark. Cells were washed and analyzed by a BD FACS Celesta (BD Biosciences, San Jose, CA) and DiVa software. Post-acquisition analyses were performed with FlowJo (TreeStar, Ashland, OR).
Titration of HBV genomes
Blood was harvested, sera were prepared, and DNA was extracted using DNeasy Blood & Tissue (Qiagen, Hilden, Germany) according to manufacturer’s protocol. The qPCR, which for each run contained serially diluted plasmids expressing the HBV sequence to provide a standard curve, was performed in a total volume of 20 µl including 1 µl of serum per PCR well. The following primers were used for amplification: forward primer: 5′-TGAGAGGCCTGTATTTCCCTGC-3′ and reverse primer 5′-AACCCCGCCTGTAACACGAG-3′. We used a ‘fast PCR’ with 40 cycles of 95 °C for 15 s followed by 60 °C for 1 min. Ct values of the standard curves were used to determine copy numbers per μl. Data were adjusted to 1 ml of serum. Water and sera from naïve mice served as controls.
Statistics
For measured continuous variables (such as T cell frequencies), two group comparisons used t-tests, provided data were normally distributed. If the normality assumption was not valid, non-parametric Wilcoxon rank-sum tests were used. Multiple comparisons were analyzed by 2-way or 1-way Analysis of variance (ANOVA) as detailed in the Figure legends.
Discussion
Most humans are able to clear an acute HBV infection, but some develop CHB. Available drugs reduce HBV replication and lower viral loads and the associated liver damage. Drugs have to be given for extended periods of time as none affects a cure and virus may resurge upon treatment discontinuation. Immunomodulators offer alternative therapies. During an acute infection innate response cytokines including type 1 IFNs, tumor necrosis factor (TNF)-α and the apolipoprotein B mRNA editing catalytic polypeptide-like (APOBEC) pathway can suppress virus replication [
25,
26]. Innate responses, to which HBV has evolved numerous escape mechanisms [
27], facilitate induction of adaptive immune responses and of those HBV-specific CD8
+ T cells can control the virus and eliminate virus-infected cells [
28,
29].
During CHV HBV-specific CD8
+ T cells lose functions [
30,
31]. HBV-specific CD8
+ T cell responses can be restored at least temporarily by adoptive transfer of T cells that are genetically modified to express a receptor against an HBV antigen. Preclinical models have yielded promising data with HBV-specific chimeric antigen receptor T cells [
32], nevertheless, due to the complexity of generating such T cells, they would not be widely available. Others have explored the use of checkpoint inhibitors to rescue exhausted HBV-specific CD8
+ T cells. Terminally exhausted CD8
+ T cells that express multiple immunoinhibitory surface markers and have undergone epigenetic changes cannot be rescued by checkpoint blockade, which mainly targets progenitor exhausted T cells [
33], which would eventually undergo transcriptional changes and differentiate into terminally exhausted T cells [
34]. In clinical trials treatment of HCC patients with a monoclonal antibody to PD1 or CTLA4 showed some, albeit limited, clinical benefits [
16,
35]. Considering that checkpoint inhibitors commonly cause serious adverse events this treatment may not be appropriate in otherwise healthy CHB patients. Vaccines that aim to induce HBV-specific CD8
+ T cells have shown efficacy in pre-clinical models especially if they were given together with an anti-PDL1 antibody [
17] but overall yielded disappointing results in clinical trials even when they were combined with anti-viral treatment or checkpoint inhibitors [
16,
36,
37].
Here we describe the immunogenicity of HBV vaccines that combine core or segments of HBV polymerase with an inhibitor of an early immune checkpoint in a pre-clinical mouse model. As we reported earlier, inhibition of the BTLA-HVEM checkpoint by HSV gD at the time of T cell activation enhances and broadens CD8
+ T cell responses [
38,
39]. gD specifically promotes CD8
+ T cells that recognise subdominant epitopes and these T cells are less prone to become impaired during chronic antigen exposure [
39]. As gD is encoded by the vaccine and largely remains localized at the site of injection and draining lymph nodes, adverse events beyond those typically observed after vaccination with an Ad vector [
40] are unlikely. Our results show that the vaccines, especially those expressing the N-terminus of polymerase, induce potent and sustained CD8
+ T cell responses to multiple epitopes that can be boosted by a heterologous AdC vector. In our current study, mice with high levels of HBV upon injection of the AAV8-1.3HBV vector generate lower frequencies of functional HBV-specific CD8
+ T cells upon vaccination and in addition exhibit a shift in the CD8
+ T cells’ epitope profile, as has been observed by others in chronic infection mouse models [
41,
42]. This raises the question if therapeutic HBV vaccines should focus on inserts that induce more modest T cell responses to subdominant epitopes, rather than inserts that are plentiful in dominant epitopes; the latter may induce more potent CD8
+ T cell responses in HBV-naïve individuals. On the other hand, vaccines rich in subdominant epitopes may fare better in patients with CHB as they may induce HBV-specific naïve CD8
+ T cells that had not been stimulated by the virus and that had therefore not been subjected to the immunosuppressive conditions of a chronic virus infection.
The AAV8-1.3HBV model is a surrogate for CHB that does not completely mirror the complexity of this disease in humans. The AAV8-1.3HBV vector, which persists in the liver of mice, does not cause any overt liver pathology and results if given at doses ≥ 10
11 vg in a few foci of infiltrating lymphocytes (data not shown). Mice injected i.v. with AAV8-1.3HBV did not develop detectable CD8
+ T cell responses to polymerase and although the vector affected the magnitude of CD8
+ T cell responses to subsequent vaccination, there was limited evidence that T cells were driven towards exhaustion; levels of exhaustion markers, including Eomes, which has recently been linked to exhaustion [
43], were not elevated on or in vaccine-induced CD8
+ T cells in AAV8-1.3HBV treated mice. The only marker suggestive of T cell dysfunctions was T-bet, which slightly decreased in the presence of HBV. The early reduction in vaccine-insert-specific CD8
+ T cell frequencies is suggestive of loss of activation rather than antigen-driven CD8
+ T cell exhaustion and remains to be investigated in more depth.
AAV vectors given to liver have been shown previously to induce transgene product-specific B and T cell tolerance due to stimulation of Tregs [
44,
45]. We assume that the reduced CD8
+ T cell responses to the Ad-gDPolN vaccine in mice that were pre-treated with the AAV8-1.3HBV vector may reflect that actions of Tregs. This is also compatible with our finding of the preferential loss of stimulation of CD8
+ T cells to immunodominant epitopes within the vaccine insert as previous work has shown them to be particularly susceptible to Treg-mediated inhibition [
46].
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.