Nanoparticle-Based Secretory Granules Induce a Specific and Long-Lasting Immune Response through Prolonged Antigen Release

1. Introduction

Protein materials, that is, supramolecular protein complexes with defined physicochemical and biological properties, are gaining interest in biomedicine because of their potent applications in drug delivery and in regenerative medicine [1,2,3,4,5]. Several approaches allow the controlled oligomerization of selected polypeptides into fibrils, layers, matrices, nanoparticles or microparticles [6,7,8,9]. One of the most versatile, promising and technologically simple strategy for protein assembly is the exploitation of the coordination capabilities between divalent cations (such as Zn²⁺, Ca ²⁺, Mg ²⁺ and Mn2+) and histidine (His) residues [10]. This category of interactivity allows the cross-molecular binding of poly-His-tagged proteins into nanoparticles and progressively [11], as granular microscale particles [12], in a process that can be reversed by the exposure to chelating agents or, more slowly, by the mere physiological, equilibrium-based dilution of the gluing cation (Figure 1A). This fact results in the progressive disintegration of the material into their forming building blocks, either nanoparticles or their protomers, that are released to the media [13,14]. In addition, it is known that the formation of the nanoparticle building blocks is favoured if the N-terminal segment of the polypeptide is cationic [15,16]. Taking this approach and by regulating the cation/His ratio in the coordination mixture, both nanoparticles and microparticles have been generated in a controlled way through robust and highly efficient protocols that keep the folding status and functionality of the forming polypeptides [17]. While the intermediate nanoparticles have been mostly adapted as vehicles for the cell-targeted delivery of small molecular weight drugs and cytotoxic protein in oncology [18,19], microparticles show appealing properties as slow drug delivery systems [20,21,22].

Once these micron-scale materials are administered subcutaneously, they leak the forming polypeptides that reach the bloodstream [22]. Notably, if fused to specific ligands of cell surface receptors, the released protein accumulates in receptor-overexpressing target organs [22,23]. This principle, which clearly mimics the secretion process of peptide hormones in the mammalian endocrine system [24,25], has been explored for the delivery of cytotoxic proteins [22] and nanoconjugates [26] in oncology, growth factors in regenerative medicine [27] and nanobodies in neurodegenerative diseases [21]. A main advantage of this system over other slow drug delivery platforms is that the protein drug is self-contained as mechanically stable but dynamic depots, in the absence of any chemically heterogeneous scaffold. Thus, inert holding materials or matrices, that might pose compatibility issues, are not needed here to support the endocrine-like character of these protein materials.

The fact that these secretory microparticles undergo a disintegration process in vivo results in a time-prolonged release of functional, properly folded proteins, in contrast to a conventional single shot. The leakage profile is dependent on protein properties [14] and on the used gluing cation [23]. This concept might be highly appealing in vaccinology as a way to expose a given antigen to the immune system during a prolonged period upon a single administration, mimicking the immune stimulation during a natural infection. Therefore, secretory microparticles might represent a novel approach to clinical immunization based on the subcutaneous implantation of antigen-delivering protein materials. This possibility has been explored here through the preparation and immunogenicity evaluation of protein-only secretory materials, utilizing two distinct antigens and in two different experimental models. Firstly, the in vivo safety and immunogenicity of protein-only microparticles based on p30 [28], a main antigen of the African swine fever virus (ASFV) [29], were examined in pigs. This structurally complex virus [30] causes severe haemorrhagic disease in domestic pigs [31]. Being a significant global veterinary concern [32] primarily due to the lack of effective vaccine prototypes [33], it is particularly worrying in large countries such as China in which the pig industry is a strong economic supporter [34,35]. Secondly, a mouse model was employed to assess the safety and immunogenicity of secretory granules based on GFP. Additionally, to obtain deeper understanding of the immunological mechanisms triggered by these materials, the specific cytokine profile induced in vitro in primary cells was examined. The administration of these secretory microparticles and the further detailed characterization of the triggered immune responses allow envisaging such a depot-based immunization approach as suited for further exploration and development of this concept. In this regard, the obtained data have been evaluated in the context of simple, cost-effective, and efficient new generation protein-based vaccination platforms that might offer appealing properties over the current immunization methods.

2. Materials and Methods

2.1. Ethics Statement

Animal care and procedures were performed in accordance with the guidelines of the Good Experimental Practice and with the approval of the Ethics Committee on Animal Experimentation of the Generalitat de Catalunya (pig experiments project codes: CEA-OH/11580/1 and CEA-OH/10298/2; mice experiment project code: CEA-OH/11691/2).

2.2. Protein Design, Production and Purification

In-house designed synthetic gene for the modular protein RK4-P30-H6 was provided by Gene Art (Thermo Fisher), subcloned in a pET22b plasmid (Novagen). The construct contains four repetitions of RK at the N-terminal that serve as a cationic peptide to stimulate nanoparticle formation together with the poly-His tag at the C-terminal end [16]. The production of the protein was tested at 20 ºC (overnight) and 37 ºC (3 hours) with two different concentrations of IPTG (0.1 mM, 1 mM) in an Escherichia coli BL21 strain. Protein production was optimized at 20ºC overnight with 0.1 mM IPTG. After the production, cells were harvested at 5000 g for 15 minutes and the pellets were washed with PBS and stored at –80 ºC until further use. Prior to the purification, cells were resuspended in Wash Buffer (20 mM Tris-HCl pH 8, 500 mM NaCl and 10 mM imidazole) with a cOmplete Protease Inhibitor Cocktail EDTA-free tablet (Roche Diagnostics). Cells were then disrupted in an EmulsiFlex-C5 system (Avestin) for 3 rounds at approximately 7500 psi. After the disruption, the mixture was centrifuged at 15000 g for 45 minutes and the soluble fraction was retained and filtered through sterile 0.22 µm filters (Millipore). Next, the filtered soluble fraction was loaded into a HisTrap HP 5 mL column (GE Healthcare) using an ÄKTA Pure chromatography system (GE Healthcare). Elution was performed through a linear gradient of Elution Buffer (20 mM Tris-HCl pH 8, 500 mM NaCl, 500 mM imidazole). The fractions were analyzed by SDS-PAGE and Western blot (anti-His tag, Genscript) and those containing RK-P30-H6 were dialyzed thrice against a saline buffer (166 mM NaCO3H, 333 mM NaCl).

2.3. Dynamic Light Scattering

The volume size distribution of nanoparticles was determined by Dynamic Light Scattering (DLS) at 633 nm (Zetasizer Pro, Malvern). Samples were diluted in their respective buffer to a concentration of 1 mg/mL. The samples were measured in triplicate.

2.4. Microparticle Formation and Protein Release

RK4-p30-H6 and GFP-H6 (at 1 mg/mL) were mixed, individually, with a zinc chloride solution at a 200:1 divalent cation:protein ratio. The mixture was incubated at room temperature for 10 minutes and then the microparticles were recovered by centrifugation (10000g, 5 minutes). The supernatant was discarded. To assess the release of protein from these microparticles, these materials were resuspended in PBS (diluted to a final concentration of 1 mg/mL) and further incubated at 37ºC for seven days. Samples were extracted on day 0, 1, 3 and 7. Then, we loaded the samples into a polyacrylamide gel which was later transferred to a PVDF membrane. The percentage of protein released during the experiment was assessed by Western Blot using an anti-polyhistidine tag antibody. Images were processed with the Image Lab (BioRad) software.

2.5. Scanning Electron Microscopy

High-resolution images of cation-induced microparticles were obtained by field emission scanning electron microscopy (FESEM). 10 μL of each microparticle sample (0.5 mg/mL) were deposited on silicon wafers (Ted Pella Inc.) overnight and then observed without coating in a FESEM Zeiss Merlin (Zeiss) operating at 1 kV and equipped with a high-resolution secondary electron detector.

2.6. Study Design of the Pig Experiments

Landrace x Large White male piglets at 8 weeks of age at arrival were used. Pigs were housed in the experimental farm of IRTA Monells (Girona, Spain). Animals were fed ad libitum, and an acclimation period of one week was allowed before the initiation of the study. In the first experiment, animals were subcutaneously inoculated once or twice three weeks apart with 50 µg (final volume of 0.5 ml) of the soluble or protein-only-microparticles (POM) form of the ASFV protein p30 in sodium carbonate buffer (166 mM NaHCO₃, 333 mM NaCl buffer, pH 8), in the presence or absence of CAF01 (half of the volume, following manufacturer’s instructions). A group of three pigs received PBS following the same procedure as control. Blood samples were collected weekly. In the second experiment, pigs received either a high dose of 150 µg of p30 POM or PBS (control group), following the same immunization regimen than experiment 1.

2.7. Study Design of the Mice Experiment

Sixteen BALB/c mice at 7 weeks of age at arrival were used (half females and half males). Mice were allocated in cages according to sex and fed ad libitum during the experiment. Ear cuts were used to differentiate the animals. After one week of acclimation, six mice (3 females and 3 males) were subcutaneously inoculated twice 3 weeks apart with 50 µg of GFP POM, while six other (3 females and 3 males) received 5 µg of the same antigen following the same regimen. GFP POM was suspended in sodium carbonate buffer (166 mM NaHCO₃, 333 mM NaCl buffer, pH 8), and the inoculated volume was 0.3 ml. Four mice (2 females and 2 males) were used as control and received 0.3 ml of PBS in each administration. Blood for the generation of serum was taken from the facial vein before the first inoculation (SD0), and 2 and 9 weeks after the second inoculation (SD35 and SD85, respectively). Two weeks after the second immunization, three mice receiving 50 µg/dose of GFP POM, three receiving 5 µg/dose and two controls were euthanized to assess the cellular response induced by GFP POM using splenocytes. The six remaining animals were kept for seven more weeks to analyse long-term immunogenicity.

2.8. ELISA

ASFV-specific antibodies were assessed by indirect ELISA based on soluble extracts of ASFV-infected cells approved by the WOAH [36]. For porcine IgG and IgA, peroxidase-conjugated rabbit anti-pig IgG (Sigma-Aldrich, A5670) at 1/20,000 or peroxidase-conjugated goat anti pig-IgA (Invitrogen, PA1-84625) at 1/2,500 were used. For specific isotypes, mouse anti-pig IgG1 (BioRad, MCA635GA) at 1/1,000, mouse anti-pig IgG2 (BioRad, MCA636GA) at 1/1,000, followed by peroxidase-conjugated goat anti-mouse polyvalent immunoglobulins (Sigma-Aldrich, A4012) at 1/2,500 were used. In the case of GFP-specific antibodies, COSTAR 3590 Corning High binding 96-well plates (Cultek, 153590) were coated with the soluble form of the protein at 5 µg/ml in carbonate-bicarbonate buffer overnight at 4 ºC. Plates were washed 3x with PBS-Tween20 0.05% (PBS-T), and blocked for 1 hour at 37 ºC, 5 % CO2 with PBS-T 2 % BSA. Afterwards, mouse serum samples in PBS-T 2 % BSA were added and incubated for 1 hour at 37 ºC, 5 % CO2. After incubation with the tested mice serum, the corresponding antibody in PBS-T 2 % BSA was used: HRP-conjugated goat anti-mouse polyvalent immunoglobulins (Sigma-Aldrich, A4012) at 1/1,000, HRP anti-mouse IgG1 (Abcam, ab97240) at 1/1,000, HRP anti-mouse IgG2a heavy chain (Abcam, ab97245) at 1/1,000, or HRP anti-mouse IgA alpha chain (Abcam, ab97235) at 1/1,000. In both cases, soluble 3,3′,5,5′-tetramethylbenzidine (TMB, Sigma-Aldrich, T4444) was used as specific chromogenic substrate for the HRP, and the reaction was stopped with 1 N H2SO4. Plates were finally read at a wavelength of 450 nm, and average absorbance [optical density (OD) values] of duplicates are represented in the graphs.

2.9. Mouse Splenocytes Collection and Flow Cytometry

Mice splenocytes were obtained by mechanically dissociation of spleens and filtration through a 40 µm cell strainer. Red blood cells were then lysed using NH4Cl for 5 min at room temperature, and splenocytes were finally suspended in RPMI 1640 medium (Gibco) supplemented with 10% FBS (Cultek), 2 mM L-glutamine (Invitrogen), 100 IU/ml penicillin/streptomycin (Invitrogen), 0.05 mM β-mercaptoethanol and 1 mM sodium pyruvate. Fresh mouse splenocytes were used for flow cytometry analysis. One million cells were used per condition in U-bottom 96-well plates. Splenocytes were stimulated for 5 days with 5 µg/ml of GFP POM. Complete RPMI was used as negative control, while stimulation with phorbol myristate acetate (PMA) plus ionomycin (at 5 ng/ml and 500 ng/ml, respectively) was used as positive control. After stimulation, cells were stained with Zombie NIR fixable viability kit (Biolegend, 423106) following manufacturer’s instructions, and then a blockage of Fc receptors was performed with PBS 5 % FBS for 15 min on ice. Extracellular staining was performed for 20 min on ice in PBS 2 % FBS using 50 µl of a mix of the following antibodies: APC hamster anti-mouse CD3e at 1/20 dilution (BD Biosciences, #553066), PerCP-Cy5.5 rat anti-mouse CD4 at 1/300 dilution (BD Biosciences, #550954) and PE-Cy7 rat anti-mouse CD8a at 1/150 dilution (BD Biosciences, #552877). Afterwards, the BD Cytofix/Cytoperm Kit (BD Biosciences) was used according to the manufacturer’s protocol to fix and permeabilize the cells. Intracellular staining using BV421 mouse anti-Ki67 at 1/150 dilution (BD Biosciences #652411) was then performed for 30 min on ice in Perm/Wash buffer (BD Biosciences). Samples were acquired on a BD FACSAria IIu flow cytometer (BD Biosciences) and data was analyzed using FlowJo v10.7.1 software (Tree Star Inc).

2.10. Multiplex Luminex Assay

PAMs were seeded in 24-well plates (10⁶ cells/well) in RPMI 1640 medium supplemented with 10% FBS, 100 IU/ml of penicillin/streptomycin, 2 mM L-glutamine and 0.5% nystatin (sigma-MERCK) and incubated overnight at 37 ºC and 5 % CO₂. Afterwards, the medium was removed and replenished with fresh RPMI either with or without 5 µg/ml of GFP POM. After a 24-hour incubation at 37 ºC, 5% CO2, cell supernatants were collected and cytokine were determined using the Luminex xMAP technology and the ProcartaPlex Porcine Cytokine & Chemokine Panel 1 (ThermoFisher Scientific). Cytokine concentrations were calculated using The xPONENT software (Luminex) and expressed as pg/ml (except for TNF, for which a technical problem with the standard curve occurred and only the mean fluorescence intensity (MFI) values could be assessed).

2.11. Statistical Analyses

Prism version 8.3.0. software (GraphPad) was used to plot the results and perform statistical analyses. The tests used are specified on figure legends. Statistical significance was set at p < 0.05 and is displayed in GraphPad style (p > 0.05 ns, * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001).

3. Results

3.1. Design and Construction of an ASFV-Antigen Nanoparticle

A full-length, proteolytically stable version of the antigen p30 was engineered and produced in Escherichia coli as a multidomain protein, placed between a cationic N-terminal peptide (RK4) and a C-terminal hexahistidine (H6) tail (RK4-P30-H6, Figure 1B). The combination of these end terminal peptides was expected to promote the self-assembling of the construct as oligomeric nanoparticles [16] assisted by the low amount of divalent cations from the media [15]. A simpler GFP-H6 construct [37] was used as a control for some experiments (Figure 1B). Upon bioproduction, RK4-P30-H6 was recovered, in a single step, by immobilized metal affinity chromatography (IMAC). The size and integrity of the construct were confirmed through Western Blot (Figure 1C, left). Notably, we observed that, as predicted, the protein spontaneously organized as monodisperse nanoparticle populations of around 52 nanometers in size, which could be disassembled by submitting the materials to denaturing plus chelating conditions (Figure 1C, right). In contrast, the control GFP-H6, also well produced (Figure 1D, left), remained unassembled because of the non-cationic character of its N-terminal end (Figure 1D, right).

3.2. RK4-p30-H6 Microparticles Promote Consistent and Prolonged Antigen Release

Using pure solutions of RK4-p30-H6 and GFP-H6, we generated secretory microscale granules by mixing the stored pure protein with a zinc chloride solution at a 200:1 divalent cation:protein ratio. The formation of these granules occurs through the Zn-mediated cross-molecular clustering [10] via the overhanging H6 tails [12], and it results in mechanically stable particles that leak the forming protein, in vitro and in vivo, during prolonged time periods [13,20,21,22,27]. The size of the resulting materials, named p30- or GFP-POM (from Protein-Only Microparticles), was determined by both SEM and DLS (Figure 2A,B), resulting in values ranging from 0.5 µm to 4 µm. We also evaluated the protein released from these POMs linked to their slow disintegration, in vitro, under physiological conditions, for seven days. Under this experimental setting, a highly regular and sustained protein leakage was observed during the entire experiment, in the case of p30-POM (Figure 2C) but faster for a large fraction of the GFP-POM content (Figure 2D). A small fraction of GFP-H6 was, however, released progressively for a few days. The differential leakage pattern of GFP-H6 was probably related to the missing N-terminal cationic peptide, which in this type of modular protein constructions shown architectonic roles through electrostatic interactions between protein monomers [16,38]. The size of the released RK4-p30-H6 and GFP-H6 proteins was compatible with their dimeric forms (Figure 2, C-D), that appeared as highly stable. This fact indicated that the building block nanoparticles (sizing around 50 nm) out of which the p30 POMs were formed (Figure 1A) were unstable upon their transit through POMs, disassembling in lower order structures in parallel or immediately after leakage in vitro. In other parallel platforms based on different proteins of oncological interest, the leaked nanoparticles have been shown more stable under in vitro testing [22]. However, we assumed that the material produced here might be more robust in vivo in a more crowded ionic environment, and that the nanoparticle version of the antigen might be at least temporarily available for immune stimulation.

3.3. Subcutaneous Administration of p30 POM Is Safe and Immunogenic in Pigs

In previous studies, the subcutaneous inoculation of protein-only secretory microparticles formed by therapeutic proteins did not show any side toxicities while the released protein was fully functional [22,23,39,40]. Here we aimed to assess the safety and immunogenicity of protein-only secretory microparticles for their potential use as slow-release systems for antigens. To do so, the soluble and POM forms of the ASFV p30 were subcutaneously administered to pigs, and the induced antibody response was analysed. In each group, pigs received one dose of 50 µg at SD0, and a booster was administered at SD14 only in half of them. For comparative purposes, two additional groups of pigs received the same protein versions but formulated with the commercial adjuvant CAF01. No adverse effects were noticed in the animals after administration of any of the formulations. Neither control pigs injected with PBS nor pigs receiving one single dose of any of the tested protein versions and formulations showed detectable ASFV-specific IgGs at any of the tested time points (Figure 3, A-E). In contrast, administration of two doses of p30 resulted in seroconversion in all cases, detectable from day 7 after the second inoculation, peaking after 14 days, and slowly declining by day 21 (Figure 3, A-D).

In previous characterization studies of secretory granules, 0.5-1 mg/doses were routinely administered to mouse models with a good accumulation of the tumor-targeted released protein in target tissues [22,23]. Considering the small amount of protein used here (50 µg/dose), we wondered whether increasing the injected POM amount could optimize the antigen formulation, and even trigger a potent immune response that might prevent the booster or the use of adjuvant. To test this hypothesis, six pigs were inoculated twice with 150 µg of p30 POM, while another group of six control pigs received PBS. Again, no seroconversion was observed after one single shot, but detectable ASFV-specific IgGs were observed after a second administration (Figure 3F). Remarkably, two weeks after the second administration (SD35), the ASFV-specific IgG titres showed that immunization with soluble p30 alone did not render significant levels compared to the control group, while 50 µg of soluble p30+CAF01 showed a slight but significant increase of specific antibodies compared to the control group at dilution 1/40 (Figure 3, G-H). Notably, the use of p30 POMV without CAF01 also resulted in a significant increase of the antibody levels at this dilution (Figure 3, G-H), thus suggesting that the particulate nature of the antigen was capable to enhance vaccine immunogenicity without the need of an adjuvant. Indeed, while the inoculation of CAF01with 50 µg of p30 POM resulted in more significantly higher levels of ASFV-specific IgGs at both tested dilutions, 150 µg of p30 POM without CAF01 rendered even more significantly higher ASFV-specific antibody levels (Figure 3, G-H).

To further characterize the antibody response triggered by p30 POM, the induced IgG isotypes as well as the presence of specific IgA in serum were assessed by ELISA. Immunization with p30, both as soluble protein version and in the POM form, induced IgG2 anti-ASFV antibodies but no detectable IgG1 (Figure 4A), indicating a Th1-like bias. This total bias towards a Th1-like response is in contrast with the balanced IgG1/IgG2 profile typically observed in sera from ASF-immune pigs after vaccination with the live attenuated ASFV BA71ΔCD2, a protective vaccine prototype developed by us [41]. Again, only the groups receiving 50 µg of p30 POM+CAF01 or 150 µg of p30 POM showed statistically significant higher levels of IgG2 compared to the controls, and more uniform levels were found in animals inoculated with the high dose of p30 POM (Figure 4A).

Notably, significant levels of ASFV-specific IgA were only found in serum after administration of the higher 150 µg p30 POM amount (Figure 4B), stressing the importance of an adequate dose.

3.4. POM Proteins Stimulate the Immune System in a Non-Specific Manner

As showed above, the molecular architecture of POMs might has an immunostimulatory capability that overcomes the lack of adjuvant (Figure 4G). This is likely due to the capacity of particulate antigens to enhance vaccine immunogenicity compared to soluble antigens [42]. Thus, to gain further insight on the immunomodulatory properties of POMs, pulmonary alveolar macrophages (PAMs) were in vitro stimulated for 24 hours with GFP POM or left untreated (RPMI), and a multiplex Luminex assay was used to assess levels of cytokines in supernatants. Our results showed that GFP POM was able to significantly stimulate the secretion of several pro-inflammatory cytokines, including TNF, IFNγ, IL-1β and IL-4 (Figure 5). No deleterious effect was observed on PAMs after in vitro stimulation with POM, supporting their safety profile for future applications.

3.5. Subcutaneous Administration of GFP POM Is Safe and Immunogenic in Mice

With the intention of compiling more data regarding the immunogenicity and the immune response triggered by POM formulations, our next step was to test them in another animal model, specifically in mice. To do so, BALB/c mice were inoculated twice 2 weeks apart with either 5 or 50 µg of GFP POM. Administration of GFP POM in mice induced long-term GFP-specific antibodies in a dose dependent manner. Thus, only the 50 µg high dose induced significant levels of GFP-specific antibodies that were detectable two weeks after the second shot (SD35, first time point tested) and lasted at least nine weeks (SD85, last time point tested) (Figure 6, A-B). The POM formulation of GFP induced significantly higher levels of GFP-specific IgG2 antibodies (Figure 6C), in accordance with what was observed in pigs receiving p30 POM. Moreover, in this case levels of IgG2 were also augmented, although not significantly, in mice receiving the high dose of 50 µg of GFP POM (Figure 6C). Also, in line with what was observed in pigs, anti-GFP IgA was detected in mice serum only when prime-boosting with 50 µg of GFP POM, but not with 5 µg doses (Figure 6D). These results confirmed the potential of POM formulations to induce mucosal immunity, as well as the dose dependence of the immune response induced. Two weeks after the second administration, half of the mice were sacrificed to analyze the cellular response induced by GFP POM. Thus, splenocytes were isolated and in vitro stimulated for 5 days with 5 µg/ml of GFP POM to be assayed for lymphoproliferative activity by flow cytometry, using the Ki67 proliferation marker. In vitro stimulation of splenocytes obtained from mice immunized with GFP POM, either with 5 or 50 µg/dose, resulted in the specific proliferation of CD8⁺ T cells (Figure 6E). Despite the proliferation extent being not significantly higher, this data provides the first evidence that GFP POM can induce memory CD8⁺ T cells in mice, even at low concentrations of antigen.

4. Discussion

In the context of global health and facing potential pandemic treats, developing new and more efficient immunization systems is an inexcusable need. Much has been learned about vaccination strategies from the still ongoing Covid-19 pandemics, but a consensus has been reached about the fact that both the classical and the recently developed vaccination strategies are still far to be optimal [43,44,45,46]. A main and generic problem of the current vaccination boosts is the limited time of exposure to the antigen. In natural infections, the sensing elements of the immune system perceive the immunogens during days or weeks, while the pathogen multiplication is active. Among other studies, a recent analysis of the response against the human immunodeficiency virus (HIV) Env protein has determined the relevance of prolonged exposure to relevant antigens in the development of vaccination strategies [47], versus the conventional single antigen shot. In fact, a slow-delivery immunization strategy based on Env over 12 days dramatically improved the immunological outcomes, as such prolonged dosage mimic features of the immune response to natural infections and expands the durability and effectiveness of such response [47]. Therefore, time-sustained dosage of immunogens is now observed as a ground-breaking and highly promising alternative to the boost-based current approaches [48].

Linked to this scenario, divalent cations as Zn⁺² are excellent protein clustering agents that exploit its interactivity with His residues to crosslink His-tagged polypeptides as nanoparticles or, at higher metal concentrations, nanoparticle-secreting microparticles [10,12,13] (Figure 1A). The microparticle version of the resulting materials show an amyloidal architecture, inner organization and protein-leaking properties [20,21,22,27]. This is similar to the properties observed in the secretory granules from the mammalian endocrine system, that support the secretion of peptide hormones from such granular depots [24,25,49,50]. This category of protein-based dynamic repositories, in the synthetic version described here, had been so far adapted to the release of protein drugs in oncology [23], degenerative disorders [21] and regenerative medicine [27]. Of course, it might be very convenient for the slow-style approach to antigen release in vaccinology. In addition, compared to other sustained drug delivery systems that require holding materials, non-functional scaffolding materials [5], bioactive polypeptides, formulated as artificial secretory granules, are self-contained and self-released in absence of potentially toxic assisting containers. Apart from the time-sustained vaccination ideology [48], the removal of any drug vehicle is also a main goal in the emerging conceptual setting around nanomedical drug-delivery [51]. This of course would be of perfect applicability to immunization approaches based on selected antigens, under the umbrella of subunit vaccines in contrast to the use of the whole pathogen.

By testing here the above principles with an innovative POM formulation, based on metal-assisted, self-assembling nanoparticles (Figure 1 and Figure 2), we demonstrate the induction of a balanced immune response, including the activation of the innate immunity as well as antigen-specific antibody and cellular responses (Figure 3, Figure 4, Figure 5 and Figure 6). The activation of the innate immune system by POM most probably provides an optimal environment for the induction of antigen-specific adaptive immune responses in the moment when the POM slowly disintegrates in the form of soluble protein components, namely unstable nanoparticles that progressively disassemble, at least in vitro, until stable dimeric protein forms. Furthermore, the induction of specific IgAs after parenteral immunization in both pig and mice (Figure 4 and Figure 6), opens new expectations for using the secretory granules as a transversal platform to generate broad systemic and probably mucosal immunity in any animal species. Further studies are required to test the efficacy of POM to induce mucosal immunity upon parenteral or intranasal administration.

In vaccinology, the activation of specific CD8 T-cell responses is of high importance to achieve sterilizing immunity against intracellular pathogens. Thus, while antibodies can block the pathogens in fluids, CD8 T-cells are the only ones capable to specifically kill the infected cells, destroying also the intracellular pathogen replicating within them. In fact, subunit and inactivated vaccines are traditionally not efficient at inducing CD8+ T cells. The ability of POMVs to induce a CD8+ T cell response observed here might be linked to their original, aggregated microscale nature, which can facilitate cross-presentation and therefore the priming of CD8 T-cell responses [52,53].

Different antibody isotypes play distinct roles in antiviral immunity. Among them, IgG2 antibodies are specifically triggered during Th1-type immune responses [54] and exhibit superior capabilities in activating Fc receptor-mediated effector responses, which are crucial for resolving infections caused by intracellular pathogens. On the other hand, elevated levels of IgG1 are associated with Th2 immune responses and do not stimulate Fc receptor-mediated immune responses as effectively [55]. Both POM formulations tested in this study induced IgG2a antibodies, regardless of the animal species used. This observation aligns with the activation of CD8⁺ T cells by cross priming and the induction of a Th1 adaptive response because of the aggregated nature of the microparticles.

5. Conclusions

As summarized from the present data, the combination of several intrinsic features of nanoparticle-based POMs, described here, make them a promising platform for further development of vaccine candidates. Among them, it is important to stress that (i) the supramolecular architecture of the antigens might be protective from proteolytic degradation in vivo; (ii) the time-extended release of the protein mimics the constant antigen stimulation during a natural infection; (iii) the supramolecular structure of the antigen might act as an adjuvant and promote antigen presentation; (iv) the delivered antigens might transit from nanoparticles to dimeric forms, through intermediate oligomers during the disintegration of the depot. Such disintegration process might induce specific antibodies and T-helper responses, and also cytotoxic T-cells by cross-priming as already shown for conventional nanoparticle-based vaccines [56,57]. Together with these features, the easy and cost-effective manufacturing, as well as the safety profile due to their intrinsic molecular purity, make POMs a very promising immunization platform with a high versatility in its further adaptation for new generation vaccine formulation.