Outer membrane vesicle loaded thin films as a stable room temperature vaccine platform

The thermostability of vaccines is paramount to their efficacy, but many vaccines are highly sensitive to temperature fluctuations and depend on cold-chain systems to maintain their immunogenicity. Outer membrane vesicles (OMVs), being derived from bacteria and containing multiple protein antigens on their surface, are no exception, and must be stored frozen long term. As such, there is a need to increase the thermostability of OMVs in order to improve access to OMV-based vaccines across the globe. Furthermore, formulation of OMVs into a platform capable of being administered via the mucosa would improve immunogenicity, as the mucosa is the main site of pathogen infection for many diseases. Here, we formulated OMVs into thin films containing various excipients and evaluated the stability of the OMV structure and antigen proteins within the OMVs over 8 weeks. Thin film compositions were optimized based on film formation ability, then four formulations were chosen and tested for OMV recovery. PMAL, an amphiphilic compound often used to stabilize biomolecules, impacted the OMVs the most, degrading their structure within 4 weeks. Other film components, including polymers and sugar alcohols, did not degrade the OMVs after 8 weeks, indicating that these compounds maintain the OMV structure in the long term. Antigen proteins on OMVs remained most stable in films containing PEG and glycerol. Films were also evaluated for their dissolution in mucosa-mimicking media, with films dissolving in less than 5 min in most solvents, demonstrating applicability as a buccally administered vaccine platform.

Infectious diseases caused by mucosal pathogens continue to pose a threat to humans across the globe. In fact, lower respiratory infections are currently the fourth leading cause of death in the world (World Health Organization, 2024), with influenza alone causing an estimated 389,000 deaths annually (Paget et al., 2019). While vaccines are continually developed to combat these pathogens, major advances are being made with vaccines that are predominantly given via injection (Lavelle and Ward, 2022). One major challenge of parenteral vaccine strategies is their long-term stability, as most are temperature sensitive, with some requiring ultra-cold freezers to maintain immunogenicity (Hosangadi et al., 2021). These ultra-cold freezers have an initial high-cost barrier and require continuous maintenance and monitoring, which is unfeasible for many vaccine campaigns in developing countries, resulting in lower vaccine uptake and coverage (Hill et al., 2016; Hosangadi et al., 2021; Kartoglu and Ames, 2022). Lack of access to these systems poses a challenge to many nations with emerging economies, as vaccines are predominantly developed by higher-income countries with stable ultracold systems (Plotkin et al., 2017; Hosangadi et al., 2021). Even if these systems are present, issues with the power grid can impact the stability of vaccines (Yakum et al., 2015). An alternative to relying on cold-chain systems to aid in the distribution of vaccines is improving their long-term thermostability. For example, it is predicted that if the thermostability of the Sanofi Pasteur pentavalent vaccine (immunizing against Diphtheria, Tetanus, Pertussis, Hib, and Hepatitis B) were improved through reformulation, it would become available in almost every part of developing countries (Lee et al., 2012). This creates a need for a more thermostable vaccine formulation that will not only be used in the United States, but worldwide.

Even more concerning, systemic administration of vaccines does not often promote immunity at the site of infection, and can often result in reduced T-cell-mediated responses, impacting vaccine efficacy (Correa et al., 2022). For respiratory infections such as influenza A, the primary site of infection is the oral and nasal mucosa, but few vaccines target administration to these locations, despite their efficacy in inducing both a local and systemic immune response (Belyakov and Ahlers, 2009; Hervé et al., 2019). Thin films provide a promising alternative delivery method for vaccines compared to injections. These platforms have been shown to preserve both active and inactive viruses, bacteria, and proteins without requiring refrigeration (Bajrovic et al., 2020; Stinson et al., 2020). Furthermore, they can be reconstituted for injection (Stinson et al., 2020), or even administered to the mucosa directly to promote a viable immune response (Bajrovic et al., 2020). Thin films have also been shown to have up to a 100% recovery of infectious and immunogenic agents depending on the film properties (Bajrovic et al., 2020).

While thin films have been previously evaluated for loading and release of both inactivated and activated viruses, outer membrane vesicles (OMVs), spherical lipid bilayers that are produced by bacteria, have yet to be investigated. OMVs have been used for the delivery of multiple antigens, including Matrix two protein (M2e), an extracellular domain of influenza A (Rappazzo et al., 2016; Watkins and Pagan, 2017; Watkins and Rappazzo, 2017; Kannan et al., 2024). OMVs are approximately 50–150 nm in diameter, are unable to replicate, and contain various pathogen-associated molecular patterns (PAMPs) that help induce a robust immune response. The thermostability of OMVs is low at elevated temperatures (above 37 °C) and they should be stored frozen or freeze-dried long term (Arigita et al., 2004), making them a strong candidate to evaluate in a thin film platform. From our own observations, precipitation of OMVs from solution can be observed within hours at room temperature, while storage at 4 °C leads to overnight precipitation and loss of protein content. Here, we investigated whether OMVs could be stably formulated into a thin film and evaluated the properties and release kinetics of OMVs from these films for their potential use as a mucosal vaccine platform.

Materials

PMAL (Poly(maleic anhydride-alt-1-octadecene) substituted with 3-(dimethylamino) propylamine) was obtained from Anatrace (Maumee, OH). DMEM, the Pierce BCA Protein Assay Kit, and Nalgene filters were obtained from Thermo Fisher Scientific (Waltham, MA). PVDF membrane and Clarity Western ECL Substrate were purchased from Bio-Rad (Hercules, CA). Coomassie blue (R-250) stain was purchased from G Biosciences (St. Louis, MO).

All other materials were obtained from Sigma Aldrich unless otherwise indicated.

OMV production, purification, and characterization

Outer membrane vesicles (OMVs) displaying four heterologous M2e epitope sequences (ClyA-M2e-4x-het containing a C-terminal 6-histidine tag) were generated and purified following previously published methods (Rappazzo et al., 2016). ClearColi Δnlp1 (CC) cells were transformed with the pBAD18 plasmid encoding ClyA-M2e-4x-het (Clya-m2e) and cultured in Luria-Bertani (LB) broth (BD Biosciences) supplemented with 50 μg/mL kanamycin at 37 °C with agitation at 225 RPM. When necessary, chloramphenicol was included at a final concentration of 50 μg/mL. Overnight cultures were diluted to an OD 600 of approximately 0.08 and grown to mid-logarithmic phase (OD 600 ∼0.4–0.6), at which point protein expression was induced with 0.2% L-arabinose. Following approximately 18 h of post-induction growth, bacterial cells were pelleted by centrifugation at 3500 RPM for 15 min. The resulting supernatant was vacuum-filtered through a 0.2 µm Nalgene filter. OMVs were subsequently isolated by ultracentrifugation using an SW 28 swinging bucket rotor in an Optima XE-90 Ultracentrifuge (Beckman Coulter) at 26,000 RPM for 3 h at 4 °C. Pelleted vesicles were resuspended in 800 µL PBS (pH 7.4).

Total OMV protein concentration was quantified using a Pierce BCA Protein Assay Kit according to the manufacturer’s instructions. Total protein content was used during formulation to ensure similar OMV concentrations across films. To verify OMV integrity, particle size distribution and concentration were analyzed by dynamic light scattering (DLS) using a Malvern Zetasizer Ultra. Analysis of particle concentration extracted from DLS and total protein content confirmed if the resuspension step yielded protein-containing vesicles.

Thin film preparation

Films were prepared via drop casting (750 uL) into 12-well plates with various solvents (dH 2 O, PBS pH 7.4, 20 mM NaCl pH 7.4, 20 mM Tris pH 7.4, and 20 mM Tris pH 8.8) to evaluate the effect of osmolarity, pH, and ion type on film drying and structure. Other excipients were added, including hydroxymethylcellulose (HPMC; 10k MW), poly(vinyl) alcohol (PVA), poly(ethylene) glycol (PEG), glycerol, sorbitol, and PMAL, at varying concentrations before addition of the OMVs. OMVs were added for a final total protein concentration upon reconstitution of 0.03 mg/mL to ensure accurate particle tracking by dynamic light scattering (DLS) and protein detection by blotting of reconstituted films. Non-loaded films were used as a negative control. Films were pipetted into 12-well plates and were allowed to dry for at least 24 h at room temperature. To assess long-term viability, films were stored for up to 8 weeks at room temperature.

Characterization of thin film properties

Scanning electron micrographs were obtained using a Zeiss EVO-50XVP with an energy of 5 kV. The films were prepared by mounting them flat on pin stub holders with double-sided carbon tape. Zoom and focus were adjusted accordingly for different measurements. In order to estimate the cross-sectional length of the films, the thin films were cut in half and then folded to image the cross-section.

Fourier transform infrared (FT-IR) spectra were obtained using a Nicolet Summit FTIR Spectrometer (Thermo Fisher Scientific) in transmission mode. The thin film samples were mounted directly on the beam path. Data were collected from 4000 to 500 cm -1 with a 4 cm -1 resolution, averaging 16 scans.

OMV recovery studies

To evaluate OMV and protein stability, films with and without OMVs were stored for varying time periods (1 day, 2 weeks, 4 weeks, and 8 weeks) covered at room temperature after drying. After storage, films were reconstituted in 3 mL of PBS each. The PBS was pipetted into each well containing the films, and the samples were allowed to dissolve for 20 min atop a rocker to provide gentle mixing. After dissolution, samples were transferred to 15 mL conical tubes and evaluated for particle count or protein content. Particle counts for each sample were collected via DLS using a Malvern Zetasizer Ultra, then the geometric average was calculated between particles. Particle counts from films not containing OMVs were subtracted out as a control. Protein content was evaluated using a dot blot. Samples were blotted (5 uL) onto nitrocellulose membrane. The membrane was blocked for 30 min at room temperature (20 °C on a rocker in 5% milk in TBS-T (pH 7.4). After blocking, the blots were probed for the histidine tag using mouse anti-His primary antibody (clone HIS.H8, 1:3000) and incubated for 90 min at room temperature. The membranes were then washed three times in TBS-T, and secondary antibody (goat anti-mouse HRP-conjugated secondary antibody, 1:5000) was added and incubated for 45 min at room temperature. Membranes were washed again three times in TBS-T, then developed using Clarity Western ECL Substrate. Membranes were imaged using a ChemiDoc MP Imaging System (Bio-Rad). The imaging protocol was set to 37 s of chemiluminescent exposure to be consistent between blots.

Protein content was also evaluated using SDS PAGE and Western blot. Briefly, samples were mixed with 4x Laemmli sample buffer (Bio-Rad) containing 2.5% β-mercaptoethanol and boiled for 3 min, then gels (12% acrylamide) were loaded and run at 65 V for 25 min followed by 110 V for 70 min. For SDS-PAGE, gels were stained with Coomassie blue R-250 (G Biosciences) and destained overnight in dH 2 O. For Western blotting, proteins were transferred to a PVDF membrane for 60 min at 100V at 4 °C. Blots were washed in TBS-T (Tris buffered saline-Tween 20, 0.1% w/v, pH 7.4), then blocked for 1 h in 5% milk in TBS-T at room temperature. Blots were probed for the histidine tag using mouse anti-His primary antibody (1:3000 as above) overnight at 4 °C. The following day, blots were washed, incubated with goat anti-mouse HRP secondary antibody (1:10,000) for 2 h at room temperature, washed, and developed using Clarity Western ECL Substrate. Both SDS-PAGE gels and western blots were imaged as above, with the protocol set to 41 s of chemiluminescent exposure.

Thin film dissolution analysis

Thin films were analyzed for their ability to dissolve in different solutions that were meant to mimic the salivary environment. Due to the ability to only remove film formulations one and two from the 12 well plate molds, only these films were used in this analysis. Films were dissolved in combinations of α-amylase (2.5 mg/mL), lipase (500 ug/mL), and lysozyme (2.5 mg/mL) in DMEM. Dissolution in PBS was used as a comparison for maximum dissolution time. Briefly, 1 mL of each solution was added to a 1.5 mL Eppendorf tube, then the films were placed into each solution. Tubes were placed in a 37 °C water bath to more accurately replicate body temperature. Thin film dissolution was considered complete when no obvious particulates were visible.

Statistical analysis

A two-way ANOVA with multiple comparisons was performed to analyze differences between films for the dissolution studies. All statistical analyses were conducted using Graphpad Prism10.

Formulation engineering of thin films

To determine the best combination of film formers, plasticizers, and surfactants, film components were mixed together, then drop casted into 12-well plates. Multiple different solvents were evaluated for their effect on film drying and structure, including dH 2 O and salt buffers. The solvent resulting in the most consistent film formation with any excipients was 20 mM Tris pH 8.8, which was used throughout the remainder of this study. Two different molecular weights of HPMC were investigated (10 kDa and 86 kDa), as well as a range of concentrations from 1% to 10%. During testing, it was found that only the 10 kDa HPMC was viable for these studies, as 86 kDa HPMC was unable to fully dissolve in the concentrations necessary to form films. Furthermore, the lowest concentration of HPMC (1%) resulted in very thin and brittle films, whereas the higher concentrations of HPMC (5% and above) were unwieldy to handle in formulations, resulting in use of 2.5% HPMC in our films. In combination with HPMC, PVA and PEG were also investigated for incorporation into the thin films, but at consistent final concentrations of 2%. Lastly, sorbitol, glycerol, and PMAL were evaluated for their influence on film properties and OMV recovery. Sugar alcohols are generally used to stabilize proteins and cells, and amphiphilic molecules such as PMAL have been hypothesized to stabilize biomacromolecules (Bajrovic et al., 2020), although the effect of PMAL on OMVs is not known.

After testing multiple film combinations (Supplementary Table S1), five films were evaluated further, with their formulations shown in Table 1. OMVs were loaded into these four film types and left at room temperature for up to 8 weeks to evaluate both particle and protein long term stability.

Physical characterization of thin films containing OMVs

Films were evaluated using FT-IR to identify chemical interactions between the film components and the OMVs. Overall, polymers appear to improve OMV stability through enhanced hydrogen bonding interactions. HPMC was an especially valuable excipient, as -OH groups from the HPMC matrix may likely be forming stronger intermolecular hydrogen bonds with functional groups (from proteins and lipids) present on the OMV surface, resulting in broader and more intense peaks (Figure 1A). The broad peak in the ∼3500 cm -1 region also overlaps amine stretching bands, with the presence of proteins likely being the major reason for peak broadening. Lipopolysaccharides present on the outer leaflet of OMVs also contain numerous hydroxyl groups that can interact with HPMC, helping to stabilize the OMV structure. Addition of PVA and sorbitol also altered the peaks, as these molecules are hydrophilic and densely packed with aliphatic hydroxyl groups, further enhancing the hydrogen bonding between the film components and the OMVs (Figure 1B). Addition of glycerol also increases hydrogen bonding due to its hydrophilicity and low molecular weight (Figure 1C).

Representative SEM images showed that films containing OMVs prepared with HPMC, PVA, and sorbitol exhibited protrusions from the film surface at various positions (Figure 1D). Interestingly, these spherical protrusions only appeared in OMV-containing films, as control films (Figure 1E) appeared uniform in surface topography. Images of the film cross section showed a thickness of approximately 50–100 µm (Figure 1F).

Long-term particle and protein stability of OMV-containing thin films

Four film types (formulations 1–4) were stored at room temperature for up to 8 weeks, then were assessed for either particle count (Figure 2A) or protein content (Supplementary Figures S1, 2B,C). After 1 day, all of the OMVs are at a similar particle count, indicating minimal degradation (Figure 2A). By 4 weeks, however, OMVs in films containing PEG (formulation 4) have had some degradation, whereas films containing PMAL (formulation 3) are fully degraded. Films containing PVA and sorbitol (formulation 1) or PVA and glycerol (formulation 4) resulted in the most OMV stability, as particle counts were consistent in these films over 8 weeks. Protein analysis, on the other hand, indicated that while the OMV structure may be stable, proteins on OMVs may be degrading at earlier time points. Dot blot (Figure 2B) was first used to analyze protein expression on intact OMVs. Despite the particle count not varying, films containing PMAL (formulation 3) show less antigen protein (ClyA-m2e) present in the dot blot by 2 weeks compared to the other formulations (Figure 2B). Films containing glycerol (formulation 2) and PEG (formulation 4) appear to have a relatively stable ClyA-m2e protein concentration over the course of the 8 weeks (Figure 2B). There is some discrepancy in formulation 1, as the protein concentration appears to be lower at earlier time points compared to later time points. To further analyze these results, SDS-PAGE was run on the samples to evaluate protein integrity (Supplementary Figure S1). Due to the low protein concentration in each film sample (∼0.03 mg/mL), and the limited sensitivity of SDS-PAGE, the ClyA-m2e protein bands (∼46.4 kDa) were not well resolved using this technique, as they should be visible around ∼50 kDa (Supplementary Figure S1). The band visible at around 37 kDa is ClyA, which is native to E. coli. Consistent with the dot blot results, however, films containing PMAL appear to show a distinctive smear and a large amount of degraded protein at the bottom of the gel that was unable to be resolved, at times as early as 1 week after film formation. Considering the low resolution of the ClyA-m2e in SDS-PAGE gels, we also ran western blots on the reconstituted thin films containing OMVs. While the dot blot aimed to show the presence of antigen on intact OMVs, Western blotting will determine the total ClyA-m2e antigen content both within and on the surface of the OMVs, as well as any his-tagged fragments produced during degradation. Interestingly, the ClyA-m2e antigens were relatively consistent throughout the 8 weeks, with some small degradation in films containing glycerol and PMAL, which is consistent with the dot blot results.

Dissolution of thin films in an oral mucosa-mimicking environment

Films were also investigated for their dissolution times in media mimicking the mucosal environment. Previous studies have determined that α-amylase (Mandel et al., 2010), lipase (Lai et al., 2019), and lysozyme (Jenzano et al., 1986) impact dissolution in the salivary environment. Due to the prevalence of these enzymes in the oral mucosa, films were tested for their dissolution times in solutions made using each of these enzymes, as well as mixtures of these enzymes (Figure 3). As formulations one and two were the only films mechanically stable enough to be fully removed from the casting molds, these films were the only ones investigated for dissolution. The results showed that there was no significant difference between the two formulations regarding dissolution time in any solvent, and that all dissolution times were below around 400 s. However, there is a visible difference in dissolution time between enzymes, with α-amylase resulting in the lowest dissolution time of the three single enzymes, as well as when present in combination with other enzymes.

Traditional routes of vaccination, including injection and intranasal administration, are highly effective strategies for global vaccination, but often do not take into account the cold-chain systems necessary for maintaining proper antigen stability. As OMVs are unstable at warmer temperatures, and there is some question as to whether the freeze-drying process damages the phospholipid bilayer (Trenkenschuh et al., 2022), OMVs are no exception to the necessity of the cold-chain system. Therefore, a platform capable of maintaining OMV thermostability is a necessary component of globalizing this type of nanovaccine for low- and middle-income countries. Furthermore, engineering a platform capable of mucosal administration would provide an opportunity for both administration at the site of infection as well as enhanced thermostability.

While many different strategies to stabilize proteins have been pursued, thin films appear to have unique potential, capable of both direct mucosal administration (Bajrovic et al., 2020) and reconstitution for use as a parenteral vaccine (Stinson et al., 2020). Additionally, thin films may be manufactured in large molds and sectioned for administration of smaller doses (Bajrovic et al., 2020). This necessitates that films can be reliably removed from their casting mold, which can prove to be a challenge depending on formulation and casting mold used. For these reasons, films that were excessively fragile or brittle were classified as unsuitable in this study. Here, we investigated multiple different compounds for their ability to form films, and determined four major film types for further analysis and OMV loading (Table 1). Of major interest were the components sorbitol, glycerol, and PMAL, as these components were included to help stabilize the OMV structure, as they have been previously used to stabilize proteins and cells (Bajrovic et al., 2020; Stinson et al., 2020). However, OMV stability could be negatively influenced by the incorporation of PMAL into the thin films due to its amphiphilic nature, as surfactants are often used to degrade OMVs in experimental assays. This was further investigated through DLS, which did indicate complete degradation of the OMVs after only 4 weeks of storage (Figure 2A).

Structural analysis indicated that the addition of polymers is favorable to OMV stability, likely through the formation of hydrogen bonding between HPMC and functional groups present on surface proteins (Figure 1A). Due to the basic pH of the solvent used in these studies (pH 8.8), lysine residues are likely deprotonated, which will result in some hydrogen bonding within the film, but not as much as if the pH was more neutral or acidic. This may enable the OMVs to interact with the HPMC strongly enough to promote stability, but not so strongly that the OMVs are being degraded. However, incorporation of glycerol in these types of films has been shown to lead to increased moisture content (Malik et al., 2022), which may result in OMV or antigen protein degradation. Indeed, there is substantial protein degradation by week 4 as shown in Figure 2B, where the ClyA-m2e protein in formulation 2 has decreased in intensity compared to formulations one and 3. Sorbitol does not seem to have the same effect (Figure 1B), which has been previously suggested to be due to sorbitol replacing water in bonding interactions and enhancing the drying process (Bajrovic et al., 2020). Overall, the higher peak intensity and peak broadening in the FTIR traces indicate that OMVs are becoming chemically integrated into the film matrix, rather than forming a separate phase or domain within the films. Visualization of the films using SEM showed physical integration of OMVs throughout the film, with protrusions present in films containing OMVs, and a less rough surface in films lacking OMVs (Figures 1D,E, respectively). It is unclear if these protrusions are the OMVs themselves or are formed by OMVs within the film matrix. As these protrusions are larger than the expected size of OMVs, the particles may be displacing the film components and forcing the formation of small air pockets, resulting in a unique surface roughness.

The size, shape, and charge of molecules are all known factors in promoting an effective immune response (Calatayud et al., 2014; Kumar et al., 2015; Fromen et al., 2016). Due to their size, structure, and the cellular components they contain, OMVs are strong adjuvants. These properties are imperative to maintain in order to promote a protective immune response, and the impact of the film components on the OMVs is a major concern. Therefore, we chose to investigate the effect of each film on OMVs in two different ways. First, we evaluated particle count using the Malvern Zetasizer, and determined the variation in the number of particles in the films over the course of 8 weeks. Due to challenges with reconstituted HPMC resulting in particles that were similarly sized to OMVs, control films were made that did not contain OMVs and were used to remove the background particle counts produced by the polymer. Furthermore, due to batch-to-batch variability, starting particle counts were taken as the particle counts on Day 1. Particle analysis showed that in formulations not containing PMAL, the OMVs were mainly stable over the course of 8 weeks, with a slight reduction in the PEG-containing film at 8 weeks (Figure 2A). As expected, the introduction of PMAL, due to its detergent-like properties, resulted in complete degradation of the OMVs after only 4 weeks of storage.

Next, to more closely inspect the antigen proteins present on the OMVs at these time points, we investigated protein integrity using dot blot (Figure 2B), SDS-PAGE (Supplementary Figure S1), and Western blotting (Figure 2C). Dot blot was used to investigate antigen protein (ClyA-m2e) present on intact OMVs as well as any degraded protein that was detectable via his tag, SDS-PAGE was used to analyze total protein content and integrity in each film sample, while Western blot was expected to show all ClyA-m2e protein present in each sample. According to the dot blots, protein degradation appears to be occurring sooner than OMV degradation in films containing PMAL (formulation 3), which is expected, as DLS may be detecting intact (or mostly intact) OMVs, but proteins on the surface of the OMVs are likely directly interacting with the film components, causing degradation to occur. This variability in target protein content in the dot blot could be due to OMV degradation. Particle degradation may result in release of additional proteins or protein fragments that are recognized by the anti-his antibody, resulting in greater signal for formulations containing these particles at later time points. Interestingly, PMAL appears to be breaking down the OMVs as early as 4 weeks, but SDS-PAGE analysis indicated that proteins appear to be degrading much earlier, as soon as 1 week after film formation, as indicated by the large amount of unresolved protein present at the bottom of the gel (Supplementary Figure S1). Due to the low concentration of antigen protein and the sensitivity limitations of SDS-PAGE, the expected bands at ∼50 kDa were unable to be clearly resolved. To investigate this further, Western blot was used to analyze the antigen content of the films (Figure 3C). The antigen content appears to be maintained across the course of the 8 weeks for films containing sorbitol (formulation 1), and to some extent PEG (formulation 4), whereas loss of antigen content is visible in PMAL-containing films (formulation 3) at 2 weeks. There is also some reduction in ClyA-m2e in glycerol-containing films at 8 weeks. The PEG-containing formulations seem to provide the most protein stability over the course of the 8 weeks, with sorbitol-PVA films producing some variability in protection of the antigen protein on intact vesicles (Figure 2B), is less apparent in the SDS PAGE (Supplementary Figure S1) and the Western blot (Figure 2C).

Vaccines administered via the mucosal (or buccal) route are of great value due to the ability of these vaccines to induce both a local and systemic immune response (Trincado et al., 2021). These types of vaccines also enable greater patient compliance by removing the use of needles (Hervé et al., 2019). In animal studies, thin film vaccines have been shown to be successfully administered via the buccal route, resulting in a similar immune response to that of a vaccine administered through the intramuscular route (Bajrovic et al., 2020). Here, we investigated how film formulations one and two dissolve in mucosa-like solvents to demonstrate applicability for future use as a buccally administered vaccine platform. Formulations one and two were used because these films were able to be consistently removed from the casting mold without damage. While there were no significant differences between the two formulations in dissolution time for any solvent combination, there were many differences among solvents. Regarding the three enzymes used, α-amylase resulted in the fastest dissolution times compared to lipase and lysozyme. This is unexpected, as cellulose compounds have been shown to inhibit α-amylase activity, and so this likely not due to enzymatic digestion of the film (Dhital et al., 2015). However, α-amylase has been shown to bind to the surface of cellulose, which may aid in the physical disruption of the film structure, enabling a faster dissolution time in amylase-containing solvent (Dhital et al., 2015).

There were many limitations to this study. We have shown that analysis of the OMVs based on particle count is possible, but there were many challenges using the Malvern Zetasizer Ultra, including the background scattering from particles caused by HPMC. Another option for more robust particle tracking may be the use of the NanoSight300, also made by Malvern Instruments. The NanoSight counts individual particles, and may be better suited to this type of analysis. Film casting into the 12-well plates produced 50–100 µm thick films, which is ideal for fast dissolution, but formulations three and four were often not mechanically strong enough to tolerate removal from the casting molds without tearing. Equivalent film integrity after removal from the casting mold was essential to ensure proper imaging and comparable dissolution conditions. We are currently investigating silicone molds, which have provided a greater success rate in film removal (data not shown). From a manufacturing perspective, the use of silicone molds could warrant the production of films with fewer defects and more uniformity between films.

Overall, we have shown that OMVs formulated into thin films are stable for up to 8 weeks depending on the formulation components and particle count can be used to evaluate OMV stability in films upon reconstitution. In the future, we plan to test the films in an animal model to determine if the immunogenicity of the OMVs remains consistent after long term storage at room temperature.