Antibody-Based Antigen Delivery to Dendritic Cells as a Vaccination Strategy Against Ebola Virus Disease

Abstract

Dendritic cells connect innate and adaptive immune responses. This is a particularly important immune checkpoint in the case of emerging infections against which most of the population does not have preexisting antibody immunity. In this study, we sought to test whether antibody-based delivery of Ebola virus (EBOV) antigens to dendritic cells could be used as a vaccination strategy against Ebola virus disease. Our approach was to use antibodies targeting the endocytic receptor DEC-205 present in murine and human dendritic cells, to deliver the EBOV nucleoprotein or the model antigen ovalbumin (OVA). Our findings indicate that DEC-205 targeting stimulated antigen-specific T-cell responses in mice, which resulted in protection from EBOV or recombinant EBOV-OVA challenge. An added value of this strategy was the generation of resident memory T cells. We propose that dendritic cell targeting could be used to improve T-cell responses against filoviruses, a strategy that may complement current vaccination strategies.

Ebola virus disease (EVD) is caused by infection with members of the Orthoebolavirus genus, often Ebola virus (Orthoebolavirus zairense [EBOV]) [1]. Most of the human populations living in endemic areas are immunologically naive against EBOV [2, 3], a scenario in which the successful transition between innate and adaptive immunity is key for survival [4–6].

Dendritic cells (DCs) are antigen-presenting cells that connect innate and adaptive immunity. The role of DCs in EBOV pathogenesis is not fully understood, but EBOV can infect some DC subtypes [7, 8].

Previous studies have harnessed the natural adjuvant function of DCs for vaccination, for example, via delivery of antigens to DCs in vivo using antibodies against DC surface receptors [9, 10]. One of the first recombinant antibodies developed was against DEC-205 (also termed CD205, LY75), which is an endocytic receptor expressing 10 membrane-external C-type lectin domains [11] expressed at high levels in type I conventional DCs (cDC1), a DC subset with high capacity to cross-present soluble antigens to CD8 T cells [12, 13]. Through this approach, antigens targeted to DEC-205 are both presented via major histocompatibility complex class I (MHC-I) and class II (MHC-II), which results in both CD8 and CD4 T-cell responses, respectively [14].

In this study, we sought to determine whether anti-DEC-205–based DC targeting could be used for vaccination against EVD. Our strategy was to use the EBOV nucleoprotein (NP) as antigen due to the fact that NP encodes the most T-cell–immunodominant peptides [4, 15, 16]. To better characterize T-cell responses, we also utilized ovalbumin (OVA) targeting to DCs followed by challenge with a recombinant EBOV-OVA virus. Our findings suggest that this strategy could be used in combination with antibody-based filovirus vaccines to improve the breadth of EBOV-specific T-cell responses.

MATERIALS AND METHODS

Ethics Statement

This study was performed with the approval of the Hamburg Animal welfare authorities (permit numbers N117/2018 and N011/2020).

Mice

Mice were purchased from the Jackson Laboratory. Bone marrow chimeras were generated by irradiating 6- to 8-week-old IFNαβR^−/− mice (strain B6(Cg)-Ifnar1^tm1.2Ees/J) twice at 550 rad 4 hours apart [8]. Two million bone marrow cells from B6 mice were transplanted to irradiated mice. The resultant chimeric mice are referred to in this study as wild-type (WT) → IFNAR^–/– mice.

Antibody Expression and Purification

Engineering of antibodies expressing mouse immunoglobulin G1 (IgG1) H and κ light regions against antimouse DEC205 (NLDC145) and nonreactive control immunoglobulin (isotype) was described previously [14]. NP-DEC205 was generated by cloning EBOV NP into the COOH terminus of the anti-DEC205 heavy chain plasmid. HEK293T were transfected with 10 µg each of the heavy and light chain plasmids using TransIT- LT1 reagent at a ratio of 3:1 (TransIT-LT1:DNA, Mirus Bio). Five hours later, 35 mL of fresh media containing 1% Nutridoma SP (Roche) and 1× L-glutamine was added. Five days after, the supernatant was harvested, and the cell debris was removed by centrifugation. Antibodies were precipitated using ammonium sulphate (Roth) (60% weight:volume) and desalted using PD-10 columns (Cytiva) after which they were purified using protein G fast flow Sepharose beads (Cytiva). Samples were incubated overnight with protein G beads and antibodies were eluted using 5 mL of a 0.1 M pH 3 glycine buffer (GE Healthcare) and neutralized with 500 µL of a 1 M pH 8 Tris-HCl buffer. The antibodies were concentrated and buffer exchange to phosphate-buffered saline (PBS) was done using Amicon Ultra-15 ultracel 10k centrifugal filter tubes (MW 100 000, Millipore Sigma). Antibody concentration was measured on a Nanodrop.

Identification of Target Cells

Eight-week-old female C57BL6 mice were intraperitoneally injected with 5 µg of allophycocyanin (APC)–conjugated anti-DEC-205 antibody. Thirty minutes later, mice were euthanized and spleen, lungs, and lymph nodes were processed for flow cytometry to assess the phenotype of APC⁺ cells.

Flow Cytometry and Tetramer-Based Staining of T Cells

Mouse spleens and lungs were harvested and digested using collagenase D (2 mg/mL, Roche) and DNaseI (50 µg/mL, Sigma). Red blood cell (RBC) lysis was performed by adding 5 mL of a 1:10 solution of RBC lysis buffer (Biolegend). Live cells were discriminated from dead cells using a live/dead dye (Zombie dye, Biolegend). A 1:1000 dilution of the dye in PBS was prepared and added to the cells for 20 minutes. Blocking of Fc receptors was done by adding an antimouse CD16/CD32 antibody (BD) for 15 minutes. Surface cell markers were stained using an antibody cocktail (Supplementary Table 1). Staining of extracellular markers was done on ice for 25 minutes. Cells were then fixed and permeabilized using 4% Cytofix/Cytoperm (BD Biosciences) for 30 minutes to 1 hour followed by intracellular staining. Sample acquisition was done on an LSR Fortessa (BD) or a Cytek AURORA (Cytek). Identification of OVA-derived SIINFEKL-specific CD8 T cells was done using commercial H-2K^b tetramers (Immudex).

Generation of Bone Marrow–Derived DCs

Femurs and tibiae bone ends were removed, and the bone marrow was flushed with medium using a 30-gauge insulin syringe. A single cell suspension was prepared by filtering the suspension through a 70-µM filter (Greiner Bio-One). RBC lysis was done as indicated above and cells were seeded in 6-well plates at a concentration of 1 × 10⁶ cells/ mL in the presence of 50 ng/mL of granulocyte-macrophage colony-stimulating factor (GM-CSF, Peprotech). Three days later, medium was exchanged with new medium containing 100 ng/mL of GM-CSF. Bone marrow–derived DCs were harvested 6 days later.

Ex Vivo T-Cell Proliferation Assay

B6 mice were immunized subcutaneously with 100 µg of a H-2K^b-restricted EBOV NP–derived peptide [17] (NP_44–52 YQVNNLEEI, JPT peptide technologies) in the presence of complete Freund adjuvant (Sigma) and boosted 21 days later with the same peptide in incomplete Freund adjuvant (Sigma). Fourteen days after boost, mouse spleens were harvested. Splenocytes were labeled with the CellTrace Violet proliferation dye (Thermofisher Scientific) according to the manufacturer's instructions. CellTrace Violet staining allows monitoring of the proliferation of T cells present in the splenocyte mix by flow cytometry, as the dye is progressively diluted with every generation of cell division. Splenocytes and bone marrow–derived cells were seeded in a round-bottom 96-well plate in the presence of 5 µg of anti-mDEC-205 EBOV NP (hereafter referred to as NP-DEC205) and 25 µg of anti-CD40 antibody (Bio X Cell). Control plates were incubated with isotype antibody controls plus 25 µg of anti-CD40 antibody.

Immunization

Groups of 8–13 mice were immunized twice intraperitoneally with 10 µg of NP-DEC205 or OVA-DEC205 (hereafter referred to as OVA-DEC205) with anti-CD40 in a maximum volume of 100 µL PBS. The boost was administered 28 days later. Control groups were similarly immunized with 10 µg of the isotype control antibody and anti-CD40.

Enzyme-Linked Immunosorbent Spot Assay

Enzyme-linked immunosorbent spot assay (ELISpot) was done using a BD Bioscience ELISpot kit following the manufacturer's instructions. EBOV NP peptides RVIPVYQVNLEEICQLIIQ (amino acids 39–58, MHC-I) and AGQFLSFASLFILPKLVVGEK (amino acids 147–166, MHC-II) were selected based on in silico analyses (IEDB, Bimas) and obtained from Pepscan/Biosynth. Individual EBOV NP peptides at a concentration of 100 µg/mL were added to each well. Spots were quantified using an AID ELISpot reader.

EBOV-OVA Rescue

The chicken OVA open reading frame was cloned as an additional viral transcriptional unit between the viral genes NP and VP into a plasmid containing the full-length genome of Zaire ebolavirus rec/COD/1976/Mayinga rgEBOV (GenBank accession number KF827427.1) [18]. Rescue of the novel recombinant virus was done as previously described [19, 20].

Experimental Infection

Groups of 8–13 mice were infected with 10⁴ focus-forming units (FFU) of WT EBOV, EBOV-OVA, or Sudan virus (SUDV). Viruses were diluted in 25 µL of PBS and were administered with a pipette intranasally to mice under isoflurane anesthesia. The virus doses utilized in this study were calculated in the model by evaluating morbidity and mortality in response to different virus doses (10² to 10⁵ FFU). No enhancement of morbidity or mortality were observed in doses higher than 10⁴ FFU. Mice were monitored daily after infection and blood was drawn via the tail vein at 3-day intervals. Morbidity was assessed based on body scoring and weight loss. Mice that reached a score of 3 and/or lost >20% of their initial body weight were humanely euthanized. Levels of aspartate aminotransferase (AST) in serum were measured using a Fuji Dri-chem analyzer. Viremia and virus titers in organs were determined using an immunofocus assay as previously described [20].

In Vivo T-Cell Proliferation Assay

WT → IFNAR^–/– mice were immunized twice with OVA-DEC205 plus adjuvant or isotype control plus adjuvant via intraperitoneal injection with 10 µg in a volume of 100 µL PBS. Twenty-four hours later, CD8⁺ T cells were isolated from the splenocytes of OT-I mice via negative selection of CD8⁺ T cells using a mouse CD8a⁺ T-cell isolation kit (Miltenyi Biotech, catalog number 130-104-075). Two million CD8⁺ T cells were resuspended in 25 µL PBS and adoptively transferred via retroorbital injection into recipient mice. Recipient mice were infected with EBOV-OVA and euthanized at different timepoints, and splenocytes and lungs were harvested for flow cytometric analysis.

Detection of Anti-EBOV NP Antibodies

Ninety-six–well plates (Nunc) were coated overnight at 4°C with 1 µg/mL of EBOV NP protein (MyBioSource) in PBS. Plates were blocked with a solution of 5% skim milk in PBS for 1 hour at 37°C. Dilutions of sera were prepared at 1:100 in 2.5% skim milk in 0.05% PBS-Tween (PBST). The plates were washed with PBST and the secondary antibody–horseradish peroxidase-conjugated goat antimouse IgG H + L (d1:10 000 in 2.5% milk–PBST buffer, Southern Biotech) was added. Plates were incubated for 1 hour at 37°C and washed thrice with PBST. A 1× solution of the substrate was prepared by mixing distilled water with TMB substrate (1:10) (enzyme immunoassay substrate kit, Bio-Rad). The reaction was stopped by adding 100 µL 1 M sulfuric acid and the optical density was measured at 450 nm using the Tecan Infinite 200 PRO reader.

NP Protein Alignment

Multiple sequence alignment of Marburg virus NP (CAA78114.1), EBOV NP (AGB56836.1), and SUDV NP (QDC12555.1) was performed on Uniprot.

Statistical Analysis

Statistical analyses were done using GraphPad Prism version 8 software. Comparisons were done using unpaired Student t test or Mann-Whitney U test. Differences in survival were assessed using the Mantel-Cox test. The levels of significance are represented as follows: P > .05, not significant; *P ≤ .05, **P ≤ .01, ***P ≤ .001, and ****P ≤ .0001.

RESULTS

EBOV NP Delivery to DCs and T-Cell Responses

To determine the nature of cells targeted by anti-DEC-205 antibodies in vivo, we inoculated mice with 5 µg of anti-DEC-205 antibodies conjugated with APC via the intraperitoneal route. Thirty minutes later, mice were euthanized and the spleens and lymph nodes collected. In both spleen and lymph nodes, anti-DEC-205 antibodies targeted primarily a population of CD11c⁺ CD8⁺ XCR1⁺ MHC-II^hi cells expressing CD24, consistent with lymphoid-resident cDC1 [20, 21] (Figure 1A, Supplementary Figure 1).

Next, we wanted to address whether DCs targeting using anti-DEC-205 antibodies fused to EBOV NP (NP-DEC205) resulted in the formation of EBOV NP–specific T cells. We generated a pool of NP-specific CD8 T cells by immunizing mice with a mouse MHC-I (H-2K^b) restricted NP-derived peptide, YQVNNLEEI, in the presence of complete Freund’s adjuvant. Mice were boosted at day 21 postimmunization and splenocytes were collected at day 14 after boost. CD8 T cells were isolated from splenocytes and labeled with CellTrace Violet, a cell proliferation tracker. We then incubated these CD8 T cells with cognate bone marrow–derived DCs in the presence of NP-DEC205 or isotype control. The addition of NP-DEC205 resulted in the proliferation of NP-experienced CD8 T cells. As a control, OVA-specific CD8 T-cell (OT-1) proliferation was similarly observed when an OVA-DEC205 fusion antibody was added to a co-culture of OT-1 and bone marrow–derived DCs (Figure 1B).

To characterize the effector functions of NP-specific CD8 and CD4 T cells, we immunized mice with either NP-DEC205 or isotype control and collected splenocytes 14 days and 28 days after. Splenocytes were pulsed with mouse MHC-II or MHC-I NP-derived peptides and subjected to interferon gamma (IFN-γ) ELISpot analysis. Both at day 14 and day 28 postimmunization, NP-DEC205 immunization resulted in the generation of NP-specific CD8 and CD4 T cells that produced IFN-γ after restimulation (Figure 1C and 1D). Our findings indicated that delivery of EBOV NP to DCs via DEC-205 targeting resulted in the production of NP-specific effector CD4 and CD8 T cells.

Anti-DEC-205 Targeting Results in Protection Against EBOV Challenge

We have previously shown that WT → IFNAR^–/– chimeras, that is, IFN-I receptor knockout mice transplanted with bone marrow progenitor cells from WT C57BL/6 mice, are susceptible to EVD [8]. Thus, we vaccinated WT → IFNAR^–/– chimeric mice with NP-DEC205 or isotype control antibody in the presence of anti-CD40 antibody [22]. Mice were vaccinated with 10 µg of NP-DEC205 or isotype control and boosted at day 28 after prime. Forty-two days after the first vaccination, mice were challenged with 10⁴ FFU of EBOV (Mayinga variant) (Figure 2A).

Vaccination with NP-DEC205 resulted in full protection from death caused by EBOV challenge. Conversely, isotype control–vaccinated mice showed a 50% lethality (Figure 2B). Both vaccinated and nonvaccinated mice showed weight loss as a sign of morbidity caused by EBOV infection, which suggested that NP-DEC205 did not provide sterilizing immunity (Figure 2C). However, vaccination resulted in reduced levels of serum AST in EBOV-challenged mice at day 9 postinfection as well as lower levels of viremia in comparison with nonvaccinated mice (Figure 2D and 2E). These results suggested a T-cell–mediated protective effect that resulted in enhanced survival. We also observed that NP-DEC205–vaccinated mice showed enhanced production of anti-EBOV NP antibodies after EBOV challenge, which indicated production of nonneutralizing antibodies against NP that could partially contribute to vaccine protection (Figure 2F).

NP Delivery to DCs Is Not Sufficient to Provide Cross-protection Against SUDV Challenge

The filovirus NP is conserved across filoviruses in particular at the protein core (Figure 3A). Thus, we reasoned that NP delivery to DCs via DEC-205 targeting could also cross-protect mice against infection with other ebolaviruses. To test this hypothesis, we vaccinated mice with NP-DEC205 or isotype control, followed by challenge with SUDV. Vaccination of mice with EBOV NP did not result in protection against SUDV (Figure 3B). There were no differences in mortality, overall weight loss, or virus titers at necropsies between vaccinated and isotype control–treated mice (Figure 3B–D). These results indicate that, despite the conserved nature of NP, vaccination of mice using this protein as target antigen did not result in cross-protection.

Protection Against Recombinant EBOV Expressing the OVA Model Antigen

Next, we rescued a recombinant EBOV expressing the model antigen OVA. In the context of mouse H-2K^b restricted T-cell responses, we reasoned that tracking of adoptively transferred OVA-specific T cells would allow us to characterize the quality of effector and memory T cells generated by DEC-205 vaccination. Thus, we inserted the OVA gene as an additional transcriptional unit after the NP gene, which allows high levels of OVA expression [23] (Figure 4A). Infectious EBOV-OVA induced a cytopathic effect in Vero E6 cells similarly to WT EBOV (Figure 4B).

We vaccinated WT → IFNAR^–/– chimeric mice with OVA-DEC205 or isotype control followed by challenge with recombinant EBOV-OVA. At day 1 after prime, mice were adoptively transferred with 4 × 10⁶ CD8 T cells isolated from the spleens of OT-1 mice, which encode a T-cell receptor specific for the OVA-derived H-2K^b-restricted peptide SIINFEKL [24] (Figure 4C). Mice vaccinated with OVA-DEC205 survived infection, whereas almost half of the animals that received isotype control antibody succumbed to EBOV-OVA infection (Figure 4D). The EBOV-OVA virus titers in organs at the time of necropsies were significantly lower in vaccinated mice in comparison with mice that received isotype control (Figure 4E), although no differences in viremia were observed (Figure 4F). These results further supported the idea that targeting of T-cell antigens derived from an immunodominant viral protein to DCs via DEC-205 may be used as a vaccination strategy against EBOV.

Characterization of Effector and Memory T Cells Generated by DEC-205 Targeting

To determine the kinetics and quality of effector and memory T cells generated in response to vaccination, we performed serial euthanasia experiments in mice vaccinated with OVA-DEC205 or isotype control, infused with OVA-specific T cells and challenged with EBOV-OVA at days 3, 7, and 21 after infection. OVA-specific T cells were detected at significant levels in lung and spleen 3 days after infection, suggesting rapid formation of effector CD8 T cells (Figure 5A). However, OVA-specific T cells were detectable at high levels in lung but not in spleen at days 7 and 21 after infection. These results suggested the formation of resident memory T cells (T_RM) (Figure 5B and 5C). Moreover, we evaluated the presence of CD8 T cells in the lung expressing the residency marker CD103. At days 7 and 21, we observed a significant increase of CD103⁺CD8⁺ T cells in the lung tissue of OVA-DEC205–vaccinated mice challenged with EBOV-OVA (Figure 5D). These results indicated that targeting of antigens to DCs results in the formation of effector T cells as well as tissue T_RM.

DISCUSSION

Most filovirus vaccines in development target the EBOV glycoprotein (GP) and elicit strong humoral responses, but exhibit weak to moderate T-cell responses [25, 26]. Antibody titers in survivors also diminish over time [27], which underlines the need for vaccines that can boost EBOV-specific T-cell immunity.

In this study we sought to determine whether targeting of EBOV NP to DCs could serve as a T-cell–based vaccination strategy. The choosing of NP was based on its conservation among filoviruses and the fact that, together with GP, it drives most of the T-cell responses in EVD survivors [5]. Our study follows the lead of various reports detailing the use of the antibody-mediated delivery of antigens to DCs, which resulted in robust T-cell responses [10, 22, 28, 29]. DEC-205 targeting can also be utilized to induce T-cell tolerance [21, 30], which opens another avenue for highly inflammatory diseases such as EVD.

A limitation of our study is that it is restricted to a mouse model with type I IFN–deficient nonhematopoietic cells. This model is susceptible to infection with nonadapted EBOV and shows a 50%–60% lethality [8], which gave us a relatively wide window to evaluate vaccine efficacy. It is, however, important to note that the levels of viremia are generally low in this model [8]. Since viremia is a biomarker of severity [31], rescuing these chimeric mice from death may be easier than in other animal models. In the initial description of the model, we achieved 50% lethality with a dose that was 10 times lower than that used in this study [8]. This was due to the necessity to refresh the IFNαβR^−/− mouse colony as well as plaque purification of our EBOV Mayinga stock. SUDV infection in this model resulted in moderate disease, which reduced the opportunity to assess vaccine efficacy against SUDV. The pathogenesis of SUDV in the WT → IFNAR^−/− chimera model requires further characterization, and it is possible that variations in administration routes or the use of other SUDV variants may result in enhanced lethality in this model. Nevertheless, despite the conservation between the EBOV and SUDV NP, we did not observe cross-protection. These results suggest that relevant mouse T-cell epitopes may be located in nonconserved NP regions and that the presence of cross-neutralizing anti-GP antibodies may be required for cross-protection. In agreement with this, natural or vaccine-induced immunity against EBOV did not provide protection against SUDV and vice versa [32, 33], and cross-protection has been only possible to achieve via monoclonal antibody therapy [34]. The mouse model, however, gave us the chance to engineer a novel recombinant EBOV expressing the model antigen OVA. Due to the availability of transgenic OT-1 and OT-2 mice that generate only OVA-specific CD8 and CD4 T cells, respectively, this model should facilitate future studies to investigate T-cell immunity in the context of EBOV infection.

Similarly to mice immunized with NP-DEC205, mice immunized with OVA-DEC205 were protected from infection with EBOV-OVA. This was further supported by OT-1 T-cell expansion and memory formation and also by the formation of T_RM. CD8⁺ T_RM are long-lived, capable of rapid proliferation, and produce proinflammatory cytokines [35, 36]. Indeed, CD8⁺ T_RM in the lungs have been shown to mediate protection to respiratory viruses [37].

Overall, our study indicates that the delivery of EBOV antigens to DCs may serve as a strategy to boost T-cell responses to EBOV and to generate long-lived tissue T_RM. DEC-205 targeting has also served to induce T-cell immunity in nonhuman primates [38] and humans, which provides an opportunity for additional preclinical studies. If not as a standalone option, we propose that this strategy could serve to improve the T-cell responses exerted by current and future filovirus vaccines.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Author contributions. C. O. performed all experiments and contributed to study design. B. S. B. rescued the EBOV-OVA virus. B. E.-P., J. R. P., A. B., E. V. N., M. H., S. W., O. B., E. A., and L. O. performed experiments in the Biosafety Level 4 laboratory. M. B.-M. and J. M.-G. provided technical support. Y. G., F. A., B. M., and J. I. contributed to experimental design and antigen selection and characterization. T. H. contributed to experimental design and supervision and rescue of EBOV-OVA. C. M.-F. designed and supervised the study and provided funding. C. O. and C. M. F. wrote the manuscript. All authors edited the manuscript.

Acknowledgments. We would like to acknowledge the Bernhard Nocht Institute for Tropical Medicine animal facility team for support with animal husbandry.

Financial support. This study was partially funded by the German Center for Infection Research (grant number TTU 01.702 to C. M.-F.) and the European Federation of Pharmaceutical Industries and Associations/Innovative Medicines Initiative 2 Joint Undertaking (grant number 116088, Pan-Ebola Vaccine Innovative Approach, to C. M.-F.). Additional funding was provided by intramural funds of the Friedrich Loeffler Institute as part of the VISION consortium (to B. S. B. and T. H.).

References

Author notes