Multilamellar Vaccine Particle Elicits Potent Immune Activation with Protein Antigens and Protects Mice against Ebola Virus Infection

Abstract

Recent outbreaks of emerging infectious diseases, such as Ebola virus disease (EVD), highlight the urgent need to develop effective countermeasures, including prophylactic vaccines. Subunit proteins derived from pathogens provide a safe source of antigens for vaccination, but they are often limited by their low immunogenicity. We have developed a multilamellar vaccine particle (MVP) system composed of lipid−hyaluronic acid multi-cross-linked hybrid nanoparticles for vaccination with protein antigens and demonstrate their efficacy against Ebola virus (EBOV) exposure. MVPs efficiently accumulated in dendritic cells and promote antigen processing. Mice immunized with MVPs elicited robust and long-lasting antigen-specific CD8⁺ and CD4⁺ T cell immune responses as well as humoral immunity. A single-dose vaccination with MVPs delivering EBOV glycoprotein achieved an 80% protection rate against lethal EBOV infection. These results suggest that MVPs offer a promising platform for improving recombinant protein-based vaccine approaches.

Graphical Abstract

Emerging infectious pathogens, such as Ebola virus (EBOV) and Zika virus, have caused a severe burden on public health, as evidenced by recent outbreaks.^1,2 As prophylactic vaccines have controlled or even eradicated several deadly pathogens in human history,³ successful vaccines would provide effective countermeasures against new emerging infectious pathogens. Viral vector-based EBOV vaccines, such as recombinant vesicular stomatitis virus (rVSV), have achieved promising disease protection rates in the field.^4–6 Yet, there are still concerns of pre-existing immunity, undesired reactogenicity, and safety issues associated with traditional vaccine approaches.^5–9 Molecularly defined subunit protein antigens can offer safer alternatives with less complicated manufacturing processes. However, on the other hand, soluble protein antigens are susceptible to deactivation or degradation, and they are less immunogenic than live vector vaccines. While various nanoparticle systems have been developed for delivery of subunit antigens to antigen-presenting cells (APCs) to induce adaptive immune responses,^10–12 it remains challenging to achieve robust, long-lasting, and concerted cellular and humoral immune responses with protein antigens.

In this work, we sought to address these issues by developing a lipid/biopolymer hybrid nanovaccine system that can deliver protein antigens to dendritic cells (DCs), promote antigen presentation, and generate CD8+ and CD4+ T cell as well as humoral immune responses in a concerted manner. We have previously reported a multilamellar lipid nanosystem formed by cross-linking of phospholipids within lipid vesicles using a bifunctional cross-linker and shown their utility for vaccine applications.^13–17 In a separate line of work, we have also reported a vaccine delivery vehicle formed by complexation of 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP) liposome−hyaluronic acid.¹⁸ Here, we sought to combine these two approaches by developing a lipid−polymer hybrid multilamellar vaccine particle (MVP) that is stabilized by multi-cross-linking of maleimide-displaying lipid bilayers with thiolated hyaluronic acid (HA-SH) (Figure 1. We chose hyaluronic acid (HA) as our biopolymer multi-cross-linker since HA is anionic for promoting fusion of cationic liposomes into multilamellar vesicles,¹⁸ biocompatible, and used widely in FDA-approved products.¹⁹ Here, we demonstrate that MVPs can be stably coloaded with protein antigens and mono-phosphoryl lipid A (MPLA), a Toll-like receptor (TLR) 4 agonist approved for use in vaccine products,^20,21 and that MVPs efficiently activate DCs and promote antigen processing. Using ovalbumin (OVA) as the model antigen, we show that immunizations with MVPs carrying OVA and MPLA induce robust expansion of OVA-specific CD8⁺ and CD4⁺ T cells and elicit durable cellular and humoral immune responses in vivo. To assess the efficacy of MVPs against EBOV, we performed immunization studies with MVPs carrying EBOV glycoprotein (GP), a recombinant protein antigen derived from the viral spike glycoprotein.^22–25 MVPs coloaded with GP and MPLA significantly enhanced GP-specific T cell immune responses, compared with soluble vaccination. Impressively, a single vaccine dose of MVPs protected 80% of mice against a lethal EBOV exposure. These results show that MVP is a promising delivery platform for subunit protein antigens and may serve as a potent delivery system for vaccination against emerging pathogens.

Figure 1.

(A) Schematic of multilamellar vaccine particles (MVPs) coloaded with protein antigen and MPLA. (B) Synthesis of MVPs. Lipid film is hydrated to form unilamellar liposomes containing maleimide-modified cationic lipids. Liposomes are then complexed and cross-linked by multithiolated hyaluronic acid (HA-SH), thereby forming MVPs.

RESULTS AND DISCUSSION

MVPs Stably Encapsulate and Display Protein Antigens.

We have developed multilamellar vaccine particles (Figure 1A) designed for stable presentation and delivery of protein antigens and adjuvant molecules in order to generate robust and durable humoral and cellular immune responses. Briefly, MVPs were synthesized by hydrating a lipid film in the presence of protein antigens and adjuvant molecules to form unilamellar liposomes, followed by incubation with biopolymers to fuse and cross-link lipid bilayers within lipid vesicles (Figure 1B). Specifically, a lipid film composed of DOTAP—maleimide-functionalized N-(4-carboxybenzyl)-N,N-dimethyl-2,3-bis(oleoyloxy)propan-1-aminium (DOBAQ)/1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), 2:1:1, m/m/m)—was hydrated and sonicated to form unilamellar liposomes displaying cationic, maleimide-functionalized lipid layers with an initial size of 130 ± 20 in diameter (Table S1). Resulting liposomes were then incubated with anionic HA polymer premodified with multiple pendant thiols (HA-SH) (Figure 1B). Addition of HA-SH promoted complexation of hydrated liposomes into multilamellar vesicles and formation of lipid− HA hybrid MVPs (Figure 1B). MPVs formed with MPLA added to the initial lipid film exhibited a hydrodynamic size of 230 ± 10 nm, while OVA, a model antigen, added during the initial liposome formation further increased the size of OVA/MPLA MVPs to 280 ± 40 nm in diameter with a polydispersity index of 0.18 ± 0.03 (Table S1). MVPs coloaded with EBOV GP and MPLA showed an increased particle size of 350 ± 5 nm in diameter (Table S1), possibly due to the higher molecular weight of GP antigen compared to OVA (MW of GP monomer 150 kDa; MW of OVA 43 kDa). Notably, the initial cationic surface charge of 19.5 ± 0.2 mV for unilamellar liposomes turned to −17.8 ± 0.4 mV after complexation with HA-SH (Table S1). Lamellarity measurement²⁶ showed that the fraction of lipid exposed on the external surface decreased from 0.57 ± 0.01 for unilamellar liposomes to 0.39 ± 0.02 for MVPs (Table S2), indicating multilamellar nanostructure of MVPs.

Next, we measured the encapsulation efficacy (EE%) of protein antigens in MVPs using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), followed by Coomassie blue staining. Mean EE% of OVA and GP in MVPs were ~18% and ~36%, corresponding to ~9 μg of OVA and ~15 μg of GP per mg of lipid, respectively (Table S3). Furthermore, we observed only ~10% leakage of OVA from MVPs incubated in 10% fetal bovine serum (FBS)-containing phosphate-buffered saline (PBS) over 7 days (Figure S1), and upon exposure to Triton X-100 detergent, MVPs retained antigens significantly more effectively than liposomes of similar lipid composition (Figure S2), thus indicating stability of antigen-loaded MVPs.

Since recognition and internalization of antigens by B cell receptors is the prerequisite for B cell activation and production of antigen-specific antibodies, we examined whether protein antigens were properly displayed on the surfaces of MVPs by an immunofluorescence assay (Figure 2A). Briefly, we measured the fluorescence signal of anti-OVA or anti-EBOV GP antibodies directed against surface antigens on MVPs and then normalized the fluorescence intensity to the particle retention (as measured by the recovery of 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindodicarbocyanine (DiD) embedded within MVPs as a marker for particle yield, Figure 2A). MVPs loaded with OVA were readily detected with anti-OVA IgG, showing a 6-fold increase in antibody binding compared with blank MVPs (P < 0.0001, Figure 2B). Similarly, MVPs loaded with EBOV GP showed a 2.6-fold increase in surface binding of a conformational anti-GP antibody, 13C6, that binds to a quaternary epitope within the GP₁/GP₂ glycan cap^27,28 (P < 0.01, Figure 2C, Figure S3). These results indicated that a subset of protein antigen encapsulated in MVPs was exposed on the particle surface and the surface-displayed antigens were readily accessible to configurational antibodies, which would allow for induction of antigen-specific humoral immune responses in vivo.²⁹ While we could not detect antigen displayed on MVPs under transmission electron microscopy (Figure S4), GP retrieved from MVPs exhibited higher MW compared with GP standards when analyzed by SDS-PAGE (Figure S5), suggesting conjugation of maleimide-displaying lipids with free thiols in proteins.

Figure 2.

Surface display of antigens on the surface of MVPs. (A) Experimental scheme. Protein-loaded and DiD-labeled MVPs were stained by a fluorophore-labeled, anti-OVA primary antibody or an anti-Ebola GP primary antibody and a fluorophore-labeled secondary antibody, followed by quantification of the fluorescence intensity between MVP with or without antigen. Results are presented as mean ± SEM, n = 3. **P < 0.01, ****P < 0.0001, analyzed by unpaired, two-tailed Student’s t-test.

MVPs Increase Antigen Accumulation and Processing by DCs.

We first examined MVP-mediated DC activation in vitro. Bone-marrow-derived dendritic cells (BMDCs) incubated with either soluble MPLA or MPLA-loaded MVPs increased the expression levels of CD80, a co-stimulatory marker (Figure 3A), and promoted DC secretion of TNF-α (Figure 3B), compared with BMDCs incubated with PBS or blank MVPs, showing that MVPs decorated with MPLA can activate DCs.

Figure 3.

MPLA-loaded MVPs activate DCs in vitro. (A) BMDCs were treated with soluble MPLA or MVPs with or without MPLA for 24 h, followed by the measurement of CD80 expression on BMDCs by flow cytometry. Representative histograms are shown. (B) Cell culture supernatant was also collected for the measurement of TNF-α by ELISA. Results are presented as mean ± SEM, n = 3. *P < 0.05, **P < 0.01, ***P < 0.001, analyzed by one-way ANOVA with the Bonferroni multiple comparison post-test.

Since antigen uptake and antigen process by APCs are both critical steps for the initiation of antigen-specific T cell responses,³⁰ we examined cellular uptake of OVA as a soluble or MVP form using BMDCs. After 2 h of incubation of BMDCs with OVA tagged with AF647, MVPs increased DC uptake of OVA by 10-fold, compared with soluble OVA, as shown by the mean fluorescence intensity (MFI) of OVA-AF647 among DCs and their representative flow cytometry histograms (P < 0.001, Figure 4A). After 24 h, we still observed a 4.4-fold enhancement in OVA uptake of MVPs, compared with soluble OVA (P < 0.001, Figure 4A). We next examined the impact of MVP-mediated antigen delivery on intracellular antigen processing. MVPs were coloaded with MPLA and DQ-OVA, which is OVA protein labeled with self-quenched fluorescent dyes that fluoresce upon protease-mediated degradation.³¹ Compared with the soluble mixture of DQ-OVA + MPLA, MVPs significantly enhanced the intracellular processing of DQ-OVA, as evidenced by a 4-fold increase in the DQ-OVA fluorescence measured by flow cytometric analysis after 24 h of BMDC culture (P < 0.001, Figure 4B). We further confirmed these results with confocal microscopy; whereas BMDCs pulsed with a soluble DQ-OVA + MPLA mixture displayed weak DQ-OVA fluorescence signal after 24 h, BMDCs treated with MVPs exhibited strong DQ-OVA fluorescence signal from endolysosomes and throughout the cytosol (Figure 4C). Taken together, these results suggested that MVPs increased antigen delivery to DCs and promoted antigen processing by DCs.

Figure 4.

MVP promotes antigen uptake and processing by DCs in vitro. (A) BMDCs were treated with OVA-AF647 or OVA-AF647-loaded MVPs for 2 or 24 h, followed by the measurement of antigen uptake by flow cytometry. Representative histograms are shown. (B, C) BMDCs were treated with the soluble mixture of DQ-OVA and MPLA or DQ-OVA/MPLA MVPs for 24 h, followed by the measurement of fluorescence intensity by (B) flow cytometry or (C) confocal microscopy. (C) Nuclei and lysosomes were stained with Hoechst 33342 and Lysotracker Red, respectively. Scale bars = 20 μm. Results in (A) and (B) are shown as mean ± SEM, n = 3. ***P < 0.001, analyzed by two-way (A) or one-way (B) ANOVA with the Bonferroni multiple comparison post-test.

MVPs Elicit Strong and Durable Antigen-Specific T Cell Immune Responses in Vivo.

We assessed the impact of MVP vaccination on induction of antigen-specific T cell responses. Naïve C57BL/6 mice were immunized subcutaneously at the tail base on weeks 0, 3, and 6 and examined for antigen-specific CD8⁺ and CD4⁺ T cell responses by IFN-γ⁺ ELISPOT (Figure 5A). Seven days after the first and second vaccination, we observed that OVA/MPLA MVPs elicited 20-fold and 11-fold higher levels of OVA-I specific splenic CD8⁺ T cell responses, compared with the soluble mixture of OVA and MPLA (P < 0.001, Figure 5B,C). To examine the durability of antigen-specific T cell immune responses, we administered the third vaccination and waited 4 months prior to performing the ELISPOT assay. Animals immunized with the MVP vaccine maintained significantly higher OVA-specific CD8⁺ T cell responses as demonstrated by a 4-fold greater average number of ELISPOT counts, compared with the soluble vaccine (P < 0.05, Figure 5D). We also confirmed systemic CD8⁺ T cell responses by SIINFEKL-H2K^b-tetramer staining of peripheral blood mononuclear cells (PBMCs) (Figure S6). In stark contrast, throughout the study, animals immunized with the soluble OVA + MPLA vaccine failed to induce statistically significant antigen-specific CD8⁺ T cell responses, compared with the PBS control group (Figure 5B–D).

Figure 5.

MVP vaccination elicits robust and durable antigen-specific T cell responses in vivo. (A) Immunization scheme. C57BL/6 mice were immunized subcutaneously at the tail base on weeks 0, 3, and 6 with PBS, soluble OVA and MPLA, or OVA/MPLA MVPs. Each dose contained 10 μg of OVA and 2.5 μg of MPLA/mouse. At 1 week after the (B, E) first and (C, F) second vaccination and (D, G) four months after the third vaccination, splenocytes were used in ELISPOT assays. Splenocytes were restimulated ex vivo with (B–D) OVA_257–264 or (E–G) OVA_323−339 peptide for quantification of IFN-γ⁺CD8⁺ or IFN-γ⁺CD4⁺ T cells, respectively. Results are presented as mean ± SEM, (B, C, E, F) n = 5–7, or (D, G) n = 5. Representative images of individual wells are shown. *P < 0.05, **P < 0.01, ***P < 0.001, analyzed by one-way ANOVA with the Bonferroni multiple comparison post-test.

In parallel, we examined the immunized animals for CD4⁺ T cell responses. Vaccination with OVA/MPLA MVPs elicited substantially greater OVA-II-specific CD4⁺ T cell responses, compared with the soluble vaccine group (Figure 5E,F), as shown by 82-fold and 10.6-fold higher IFN-γ⁺ ELISPOT counts after the first (P < 0.05) and second (P < 0.01) vaccination, respectively. In addition, an ELISPOT assay performed at 4 months after the third vaccination showed that MVP immunization sustained markedly stronger antigen-specific CD4⁺ T cell responses, as demonstrated by a 7.8-fold greater average number of ELISPOT counts, compared with the soluble vaccine (P < 0.05, Figure 5G). We sought to further examine induction of antigen-specific CD4⁺ T cell activation by MVPs. BMDCs pulsed in vitro with OVA/MPLA MVPs significantly enhanced the proliferation of OT-II CD4⁺ T cells with a T cell receptor (TCR) specific for the OVA_323–339 peptide (sequence: ISQAVHAAHAEINEAGR), compared with the soluble OVA + MPLA mixture (P < 0.001, Figure S7A). To validate these results in vivo, we adoptively transferred OT-II CD4⁺ T cells expressing a CD45.2⁺ congenic marker into CD45.1⁺ recipient mice on day 0, followed by vaccination on day 1 with OVA and MPLA in either soluble or MVP form (Figure S7B). Analysis of transferred CD45.2⁺ T cells in the spleen on day 8 showed that OVA/MPLA MVPs significantly induced proliferation of OVA-specific OT-II CD4⁺ T cells, compared with the soluble vaccine (P < 0.001, Figure S7C).

Taken altogether, these results demonstrated that MVP vaccination markedly improved T cell immune responses, compared with vaccination with the soluble mixture. In particular, OVA/MPLA MVPs achieved robust antigen-specific T cell responses, characterized by balanced CD8+ and CD4+ T cell responses that were maintained even after 4 months of vaccination. This highlights the potency of MVP vaccination for eliciting robust and durable T cell immune responses with subunit protein antigens.

Nanoparticle Vaccines Elicit Robust Humoral Immune Responses.

We also examined humoral immune responses induced by vaccination with OVA/MPLA MVPs. Following the immunization scheme shown in Figure 5A, we immunized animals, collected serum at week 3 after each vaccination, and quantified anti-OVA total IgG, IgG₁, and IgG_2c serum titers. Compared with soluble vaccination, OVA/MPLA MPV vaccination achieved a 19-fold higher level of anti-OVA serum IgG by 9 weeks (P < 0.001, Figure 6A) and maintained robust antibody responses throughout 18 weeks (P < 0.001, Figure 6A). Analysis of antibody subtypes indicated that compared with soluble vaccination, MVP vaccination induced 9-fold (P < 0.05) and 13-fold (P < 0.01) higher levels of anti-OVA IgG₁ serum titers by weeks 9 and 18, respectively (Figure 6B). Similarly, MVP vaccination elicited 52-fold (P < 0.001) and 69-fold (P < 0.05) higher levels of anti-OVA IgG_2c serum titers than soluble vaccination on weeks 9 and 18, respectively (Figure 6C). These data indicated that MVP vaccination generated robust, Th₁/Th₂-balanced humoral immune responses.

Figure 6.

MVP vaccination induces potent humoral immune responses in vivo. C57BL/6 mice were immunized as in Figure 5A. Sera were collected after 3 weeks of each vaccination and 3 months after the third dose. Serum samples were quantified with ELISA for measuring serum titers of anti-OVA (A) total IgG, (B) IgG₁, and (C) IgG_2c. Results are presented as mean ± SEM, n = 5. *P < 0.05, **P < 0.01, ***P < 0.001, analyzed by two-way ANOVA with the Bonferroni multiple comparison post-test.

GP/MPLA Co-loaded MVP Induced Strong Vaccine Efficacy against EBOV.

Based on the compelling immunogenicity results of MVP vaccination, we assessed the MVP platform for vaccination against EBOV. In particular, we synthesized MVPs carrying an EBOV GP antigen, which is a recombinant protein antigen derived from the EBOV viral spike glycoprotein.²² We incorporated GP and MPLA into MVPs using the same protocol as presented above and analyzed antigen-specific T cell immune responses after immunizations on days 0 and 21 (Figure 7A). In addition, we directly compared MVPs to our previously reported multilamellar lipid nanosystem formed by cross-linking of lipid layers using a bifunctional cross-linker (termed ICMVs).¹³ For assessing T cell immunity, we performed ex vivo restimulation of splenocytes with WE15 (WIPYFGPAAE-GIYTE), a GP-specific peptide with epitopes for both major histocompatibility complex (MHC)-I and -II. GP/MPLA MVPs markedly enhanced GP-specific T cell responses, as shown by 10-fold and 3.5-fold greater IFN-γ⁺ ELISPOT counts, compared with the soluble and ICMV vaccine groups, respectively (P < 0.05, Figure 7B). More specifically, intracellular cytokine staining for IFN-γ and TNF-α indicated that GP/MPLA MVP vaccination significantly increased GP-specific, polyfunctional CD8⁺ and CD4⁺ T cell responses, compared with the soluble vaccine (P < 0.05, Figure 7D–G). All vaccine groups increased GP-specific B cell frequencies, but there was no statistical difference among the vaccine groups (Figure 7C).

Figure 7.

MVP vaccination generates strong EBOV GP-specific immune responses. (A–G) C57BL/6 mice were immunized subcutaneously with soluble GP + MPLA, GP/MPLA ICMVs, or GP/MPLA MVPs on days 0 and 21. At 1 week after the final dose, splenocytes were collected and restimulated ex vivo for quantification of antigen-specific (B) IFN-γ⁺ T cells or (C) B cells by ELISPOT. (D–G) Intracellular cytokine staining was performed to quantitate the percentage of (D, F) IFN-γ+ and (E, G) TNF-α⁺ T cells among (D, E) CD8⁺ and (F, G) CD4⁺ T cells. Results are presented as mean ± SEM, (B, C) n = 3 and (D, E, F, G) n = 4, 5. *P < 0.05, **P < 0.01, ***P < 0.001, analyzed by one-way ANOVA with the Bonferroni multiple comparison post-test.

Having demonstrated immunogenicity of GP/MPLA MVPs, we examined their efficacy to protect animals against EBOV viral infection. As there is an urgent need to develop rapid yet durable countermeasures against bioterrorism agents,¹² we employed a rigorous test model where animals were challenged with a lethal dose (1000 particle forming units) of mouse-adapted EBOV on day 28 after a single-dose vaccination with 3 μg of GP (Figure 8A). A single dose of MVP vaccination significantly enhanced GP-specific IgG serum titers, achieving on average 128-fold higher serum titers compared with soluble GP + MPLA vaccination (P < 0.001, Figure 8B). Importantly, a single-dose MVP vaccination protected 80% of animals against EBOV viral challenge, whereas soluble vaccination protected only 10% of animals (P < 0.01, Figure 8C). While a single-dose ICMV vaccination also induced robust GP-specific IgG serum titers, ICMVs protected only 40% of animals (Figure 8B,C), suggesting that concerted T cell and IgG responses elicited by MVPs (Figure 7B–G, Figure 8B) may have contributed to their strong efficacy. Overall, these results demonstrated that MVP vaccination elicited protective immunity against lethal EBOV exposure.

Figure 8.

MVP vaccination protects mice against lethal EBOV infection. (A–C) C57BL/6 mice (n = 10) were immunized subcutaneously on day 0 with a single dose of GP and MPLA formulated in MVPs or ICMVs (3 μg of GP and 2.5 μg of MPLA) or as a soluble mixture (10 μg of GP + 2.5 μg of MPLA). (B) Sera collected on day 14 were assessed for GP-specific IgG serum titers by ELISA. (C) Preimmunized mice were inoculated intraperitoneally on day 28 with a lethal dose (1000 pfu) of mouse-adapted EBOV and monitored for animal survival. Results are presented as mean ± SEM, n = 10. (B) ***P < 0.001, analyzed by one-way ANOVA with the Bonferroni multiple comparison post-test. (C) **P < 0.01, ***P < 0.001, analyzed by the log-rank (Mantel–Cox) test.

In this work, we have demonstrated the general applicability of MVPs using both a model antigen OVA and an EBOV subunit antigen. In particular, immunizations with GP/MPLA MVPs elicited strong GP-specific adaptive immune responses (Figure 7), and a single MVP vaccination with a 3 μg GP dose exhibited an impressive 80% protection rate against lethal EBOV exposure given at 4 weeks postvaccination (Figure 8), thus demonstrating protection against ma-EBOV infection under dose-sparing conditions of a recombinant EBOV GP antigen. It is difficult to make direct comparisons between our EBOV study results and those of other previous EBOV vaccine strategies due to differences in vaccine schedules, modalities, and dose. Nevertheless, prior studies have reported protective immunity against EVD in rodent models with a recombinant EBOV GP fused to human IgG₁ Fc formulated with adjuvant;^32,33 however, they achieved protection after prime–boost–boost vaccination with a GP antigen dose of 25–100 μg.^32,33 Similarly, prime–boost–boost immunizations with 10 μg of EBOV GP mixed with a proprietary saponin-based TLR4 agonist or a complex squalene-based emulsion have been reported to protect mice against a lethal EBOV exposure.^24,34 However, it remains to be tested if such previous approaches would offer a single-dose protection against EBOV infection with low antigen dose as in our current MVP EBOV studies.

Despite extensive research in the field of Filoviruses, immune correlates of protection against EVD are largely unknown.^5,6 Nevertheless, detailed rodent and nonhuman primate studies suggest that both potent EBOV GP-specific humoral immunity and CD4⁺ and CD8⁺ T cell responses are required for complete protection.^35–39 On the basis of these findings and our immunogenicity data, we speculate that robust, concerted humoral and cellular immune responses elicited by MVP vaccination collectively contributed to the protective immunity against EBOV infection observed in our proof-of-concept studies. Further studies are needed to delineate the mechanisms of action for MVP vaccination.

MATERIALS AND METHODS

Materials.

Lipids including DOPC, DOTAP, and MPLA were purchased from Avanti Polar Lipids. The maleimide-modified lipid DOBAQ-MAL and HA-SH were synthesized as described previously.⁴⁰ Ovalbumin was from Worthington Biochemical Corporation, DQ-OVA was from ThermoFisher Scientific, and the Ebola glycoprotein was kindly provided by Dr. Christopher L. Cooper (USAMRIID). All other materials were at least reagent grade.

Synthesis and Characterization of MVPs.

Lipids composed of DOTAP, DOPC, and DOBAQ-MAL (1:0.5:0.5, m/m) plus MPLA were dried to form the lipid film, which was then hydrated with proteins (50 μg of OVA or 40 μg of GP/1.26 μmol of total lipids) dissolved in PBS buffer (pH 7.4) at room temperature for 1 h under intermittent vortex, followed by brief sonication, addition of HA-SH (200 μg/1.26 μmol of total lipids), and incubation at 37 °C for 1 h under 700 rpm constant shaking to promote complexation and cross-linking. The resulting particles were washed with PBS by centrifuging (20817g, 5 min) three times, dispersed by brief sonication, and stored at 4 °C before use. ICMVs coloaded with GP and MPLA were synthesized as reported previously.¹³

Particle size and surface charge were measured by dynamic laser scattering (DLS) using the Zetasizer Nano (Malvern, UK). Particle morphology was visualized by transmission electron microscopy (TEM). In brief, drops of nanoparticle solution were adsorbed for 1 min to a glow discharged 400-mesh copper grid covered with carbon-coated colloidion film (Structure Probe, Inc.). The grids were washed twice with deionized water and then negatively stained in 0.7% uranyl formate. TEM images were obtained on a FEI Morgagni electron microscope run at 100 keV and a magnification of 22 000× (2.1 Å/pixel) and then recorded on a Gatan Orius charge-coupled device camera.

The encapsulation efficiency of proteins was measured by SDS-PAGE followed by Coomassie Blue staining. The lamellarity of MVPs was measured as previously described.²⁶ In vitro leakage of the encapsulated protein antigen was measured by a fluorescence-based assay. In brief, particles loaded with Alexa Fluor 647-labeled OVA (OVA-AF647) or soluble OVA-AF647 (control) were added into a 300 kDa MWCO dialysis tube (Spectrum G235060) and incubated in PBS supplemented with 10% FBS at 37 °C under 120 rpm constant shaking. Samples outside the dialysis tube were collected at predetermined time points within 1 week and quantified for fluorescence intensity using a plate reader (Synergy Neo, BioTek, USA). The protein release was also examined under detergent treatments. Briefly, OVA-AF647-loaded MVPs and liposomes (composed of the same lipids as in MVPs) were treated by different concentrations of Triton X-100 in PBS and analyzed for protein release immediately or after incubation for 6 h at 37 °C. The percent release was calculated by dividing the fluorescence intensity of the supernatant by the total fluorescence intensity after complete disruption of particles with 1% Triton X-100 in PBS.

To determine the display of proteins on the surfaces of MVPs, OVA-loaded and DiD-labeled (0.1% mol of total lipids) particles were stained by an anti-OVA-FITC antibody (Abcam ab85584, 1:100 dilution) at room temperature for 1 h, followed by multiple washes using an airfuge (200000g, 10 min). The pellet was resuspended in 1% Triton PBS and measured for the fluorescence intensity at Ex/Em = 488/520 (FITC) nm and 650/680 nm (DiD) using a plate reader. Ebola GP-loaded particles were first stained with a primary antibody (mouse-13C6, provided by Dr. Christopher L. Cooper, USAMRIID, 1:100 dilution), followed by washes and staining with a goat antimouse IgG-PE secondary antibody (eBioscience 12-4010-87, 1:50 dilution), and measured for the fluorescence intensity as above. In a separate study, the specificity of the primary antibody was determined by staining DiD-labeled and GP-loaded MVPs with 13C6 or an isotype control antibody, followed by the same sample processing as above.

Activation of BMDCs in Vitro.

Murine bone-marrow-derived dendritic cells were prepared according to a previous protocol⁴¹ and used within days 7–11 of culture. DCs were seeded into a 24-well plate at 2 × 10⁵ cells/well and treated with soluble MPLA or MVPs with or without MPLA loading at 0.5 μg/mL MPLA for 24 h. Cells were then collected and measured for the expression of CD80 (eBioscience 12–0801) by flow cytometry (CytoFLEX, Beckman Coulter, USA). Cell culture supernatant was also collected for the measurement of DC secretion of TNF-α by ELISA.

Uptake and Intracellular Processing of Protein Antigens by BMDCs in Vitro.

For the antigen uptake study, BMDCs were seeded into a 24-well plate at 2 × 10⁵ cells/well and incubated with soluble OVA-AF647 or OVA-AF647 MVPs at 0.8 μg of protein/mL for 2 or 24 h, followed by the measurement of fluorescence intensity among CD11c⁺ (BD558079) DCs by flow cytometry. To investigate the intracellular processing of protein antigens, MVPs were loaded with MPLA and DQ-OVA (OVA labeled with a self-quenched fluorescent probe, which will fluoresce upon protein degradation) and added to the DC culture at a dose of 2 μg of DQ-OVA and 0.25 μg of MPLA/2 × 10⁵ cells, followed by an incubation for 24 h and measurement of fluorescence intensity in DCs by flow cytometry. The soluble mixture of DQ-OVA and MPLA was also included as a control. For the confocal microscopy study, DCs were seeded on glass coverslips put into a 24-well plate and incubated with the soluble mixture of DQ-OVA and MPLA or DQ-OVA/MPLA coloaded MVPs at 1 μg of DQ-OVA and 0.2 μg of MPLA/2 × 10⁵ cells for 24 h, followed by washes and staining of cells with 0.1 μM Lysotracker (ThermoFisher L7528) and 1 μg/mL Hoechst 33342 (ThermoFisher H3570) at 37 °C for 1 h. Cells were then fixed using 4% paraformaldehyde solution, coated on a glass slide using the ProLong Diamond Antifade Mountant (ThermoFisher P36965), and visualized using a confocal microscope (Nikon A1, USA).

Proliferation of Antigen-Specific CD4⁺ T Cells ex Vivo.

BMDCs were seeded into a 96-well plate at 5 × 10⁴ cells/well and treated with soluble OVA plus MPLA or OVA/MPLA coloaded MVPs at a series of concentrations for 4 h. OVA-specific CD4⁺ T cells were purified from OT-II mice (kindly provided by Dr. Lonnie Shea, University of Michigan) using a EasySep kit (STEMCELL 19765), labeled with the CFSE fluorescence dye (ThermoFisher C34570), and added to the treated BMDCs at 5 × 10⁴ cells/well. After coculture for 3 days, the proliferation of T cells was measured by dilutions of the CFSE fluorescence by flow cytometry.

Proliferation of Antigen-Specific CD4⁺ T Cells in Vivo.

CD45.1⁺ C57BL/6 mice (six-week-old, female, The Jackson Laboratory, USA) were intravenously transferred with 3 × 10⁵ naïve CD4⁺ T cells isolated from CD45.2⁺ OT-II mice. One day later, recipient mice were immunized with PBS, the soluble mixture of OVA and MPLA, or OVA/MPLA MVPs at 10 μg of OVA and 2 μg of MPLA/mouse. One week later, splenocytes from recipients were collected and stained with CD4 (BD 560569) and CD45.2 (BD 561874) for quantification of the percent of transferred CD45.2⁺ cells among total CD4⁺ splenocytes using flow cytometry.

MVP Immunization and EBOV Virus Infection Studies.

Animal studies were conducted under an Institutional Animal Care and Use Committee (IACUC) approved protocol in compliance with the Animal Welfare Act, PHS Policy, and other Federal statutes and regulations. Animal experiments employing the OVA and GP antigen were performed under approval from IACUC at the University of Michigan and USAMRIID, respectively. All animal experiments were conducted at an Association for Assessment and Accreditation of Laboratory Animal Care, international accredited facility. C57BL/6 mice were immunized subcutaneously at the tail base with PBS, the soluble mixture of OVA and MPLA, or OVA/MPLA MVPs every 3 weeks for three doses. Each dose was composed of 10 μg of OVA and 2.5 μg of MPLA. At 1 week after the first and second doses and 17 weeks after the third dose, splenocytes were collected and purified using the lymphocyte separation medium (Lonza 17–829), seeded into the IFN-γ ELISPOT plate (BD 551083) at 2 × 10⁵ cells/well, and cultured with 20 μg/mL OVA_257–264 or OVA_323–339 peptide for 18 h. IFN-γ⁺ spots were measured according to the manufacturer’s protocol (BD 551083). At 17 weeks after the third dose, blood was also collected for quantification of the percent of peripheral OVA-specific CD8⁺ T cells using a SIINFEKL-H2K^b-PE tetramer (MBL, USA).⁴² Under the same immunization scheme, sera were collected at 3 weeks after each dose and 9 weeks after the third dose for analyses of serum titers of anti-OVA total IgG, IgG₁, and IgG_2c by ELISA. Titers were designated as the reciprocal of the highest dilution factors with OD₄₅₀ values at least 2 times higher than the background values obtained from the PBS immunization group.

For studies using the EBOV GP antigen, C57BL/6 mice were immunized subcutaneously at the tail base with the soluble mixture of GP and MPLA or GP/MPLA coloaded MVPs or ICMVs every 3 weeks for two doses. Each dose was composed of 3 μg of GP and 3 μg of MPLA. Splenocytes were collected at 1 week after the second dose, restimulated ex vivo with 4 μg/mL of a GP-derived WE15 peptide (sequence: WIPYFGPAAEGIYTE), and quantified for GP-specific IFN-γ⁺ T cells or B cells by the ELISPOT assay. Restimulated splenocytes were also quantified for the percent of intracellular IFN-γ⁺ or TNF-α⁺ by flow cytometry. In another study, mice received a single vaccine dose, followed by serum collection 2 weeks later for the analysis of anti-GP IgG titers by ELISA. Immunized mice were infected intraperitoneally with 1000 pfu of a mouse-adapted Ebola virus/H.sapiens-tc/COD/1976/Yambuku-Mayinga (ma-EBOV) one month later and monitored for the animal survival. Studies conducted with ma-EBOV were under the maximum BSL-4 containment.

Statistical Analysis.

Data were analyzed by the unpaired, two-tailed t-test or the one or two-way analysis of variance (ANOVA) followed by the Bonferroni’s post-test for comparison of multiple groups using Prism 5.0 (GraphPad Software). P values less than 0.05 were considered statistically significant. All values are presented as mean ± SEM with an indicated sample size.

CONCLUSIONS

We have developed MVPs based on lipid−hyaluronic acid hybrid nanoparticles for vaccine applications. MVPs improved antigen delivery and antigen processing by DCs, generating robust and long-lasting antigen-specific CD8⁺ and CD4⁺ T cell responses as well as humoral responses. A single dose of MVP vaccination protected 80% of animals against lethal EBOV infection. MVP is a promising vaccine delivery system and warrants further investigation as a vaccine platform for recombinant protein antigens.