Cyanobacterial Extracellular Vesicles as Protein Carriers: Towards Fish Vaccination

ABSTRACT

Fish aquaculture faces significant economic losses from disease outbreaks. Vaccination is the most effective prevention strategy, and bacterial extracellular vesicles (EVs) show promise as vaccine platforms due to their strong immuno‐stimulating properties. However, the application of EVs derived from pathogenic bacteria is limited by toxicity risks and production challenges. Alternatively, genetic engineering of non‐pathogenic microorganisms is being explored to produce tailored EVs to deliver antigens and serve as carriers of therapeutic proteins. Recently, we have engineered the model cyanobacterium Synechocystis sp. PCC 6803 for the expression of the reporter green fluorescent protein (sfGFP) and its targeting to EVs. Here, taking advantage of the Synechocystis sfGFP‐loaded EVs, the stability of vesicles and their cargo was evaluated in the long term when stored under different temperature conditions and after freeze‐drying. The possibility of using Synechocystis EVs as a tool for eliciting specific/adaptive immune responses was assessed in European seabass, a high commercial value fish, by following the amount of total and sfGFP‐specific immunoglobulins produced after immunisation through injection. Synechocystis EVs were shown to be resilient nanostructures that can induce specific immune responses in fish with additional adjuvant features. This represents a biotechnological breakthrough towards a novel antigen‐carrier platform for sustainable fish‐pathogen control.

Here, we describe that Synechocystis EVs are stable nanostructures that can induce specific immune responses in fish with additional adjuvant features. It represents a biotechnological breakthrough towards a novel antigen‐carrier platform for sustainable fish‐pathogen control in aquaculture settings.

1. Introduction

The constantly growing demand for fish food has resulted in a large‐scale intensive type of aquaculture, which has been associated with increasing disease outbreaks. These occurrences severely affect the aquaculture industry's profitability, sustainability, and reputation. Several factors can impact the spread of diseases in aquaculture environments, including high stocking density, poor water quality, water temperature fluctuations, or contact with other fish. If not adequately handled, transferring infected fish or using contaminated equipment or water can spread these diseases from pond to pond or aquafarm to aquafarm. As in other food‐producing sectors, infections can be controlled by making use of chemotherapeutics, such as antibiotics, which is a strategy that raises concerns regarding drug resistance and safety, and its application is strictly limited by legal regulation. Thus, alternative strategies to control fish infections are urgently needed, and vaccination is widely accepted as the most effective and environmentally friendly solution (Dadar et al. 2017; Mackenzie and Jeggo 2019).

Vaccines have been used in fish farming for more than 40 years and represent one of the main reasons for the success of fish production (Ma et al. 2019). Most licensed products are based on live attenuated or inactivated pathogens. However, getting hold of the pathogen, growing it in sufficient amounts for subsequent attenuation or inactivation, and later producing the vaccine raises concerns about the safety risks of handling the pathogen. Moreover, its effectiveness is not always guaranteed, as pathogen‐inactivation processes denature antigenic agents, reducing immunogenicity. Other types of vaccines exist to overcome these hurdles, such as subunit vaccines (Ma et al. 2019), based on synthetically prepared, recombinant, or highly purified antigens in a suitable carrier or adjuvant. Taking advantage of the fact that they do not replicate in the host, presenting no risk of pathogenicity, subunit vaccines are becoming increasingly tested and implemented. Other advantages of subunit vaccines include their highly characterised production state and the possibility of being freeze‐dried, which enables non‐refrigerated transport and storage (Hou et al. 2023). Nevertheless, their ability to elicit a potent immune response may be weaker than that of whole‐cell preparations due to the limited number of components available capable of stimulating an immune response.

Vaccines based on inactivated pathogens or recombinant antigens are usually unable to confer protection on their own, requiring the use of adjuvants or immunostimulants to increase vaccine efficacy (Christensen 2016). Thus, the development of effective fish vaccines should consider not only the search for protective antigens, but also seek biocompatible adjuvants that maximise immunogenicity with a desired immune response. Traditionally, oil emulsions are used as adjuvants (including mineral oils and Freund's complete adjuvant). However, others have been tested, namely nano/microparticles (e.g., encapsulation of vaccines in biocompatible and biodegradable poly‐(lactide‐co‐glycolide) [PLGA] polymers), aluminium, β‐glucans, saponins, and even lipopeptides (Vinay et al. 2018; Zhao et al. 2023).

One of the success stories in European aquaculture is the European seabass ( Dicentrarchus labrax ), a species of high economic interest, especially in the Mediterranean region, where production grew from 58 to 191 ktonnes in less than 15 years (2002–2016) (European Market Observatory for Fisheries and Aquaculture Products [EUMOFA] 2018). Although considered a sturdy species, European seabass is subject to a wide range of diseases (Food and Agriculture Organization of the United Nations [FAO] 2025), namely vibriosis, photobacteriosis, and tenacibaculosis, which are significant constraints that limit European seabass productivity. Nevertheless, the number of commercially available vaccines for European seabass immunisation is still limited, as is the case for many commercially relevant aquaculture species. To further advance the field of fish vaccination, novel antigen‐delivery strategies are required (Dadar et al. 2017), which can improve antigen protection during storage and ease the administration process, combined with adequate adjuvants, towards stimulating effective fish immune responses.

Bacterial extracellular vesicles (EVs) are nanosized, spherical particles with a diameter ranging between 20 and 400 nm, usually released by the cell envelope of bacteria (Lima et al. 2020). They are formed by membrane components (lipids, lipopolysaccharides in the case of Gram‐negative bacteria, and proteins), surrounding a lumen with soluble content (proteins, metabolites, and even nucleic acids) (Schwechheimer and Kuehn 2015; Woith et al. 2019; Lima et al. 2020). Bacterial EVs have already been used in vaccinology (Zhu et al. 2021), a strategy that has followed essentially two main approaches. On the one hand, EVs naturally released from the pathogenic bacterium are isolated and directly administered (van der Pol et al. 2015; Lieberman 2022). In this case, EVs contain several pathogen‐associated molecular patterns that trigger an immune response in the target group, the best example being the commercially available vaccine Bexsero (Deghmane and Taha 2022). On the other hand, EVs derived from non‐pathogenic bacteria can also be used, an approach that explores the capacity to genetically engineer model bacteria to produce recombinant EVs loaded with heterologous antigen(s) (Kim et al. 2008; Valderrama and Gutierrez 2018). As the latter offers great versatility and flexibility, it has been increasingly implemented in various target groups. In both approaches, however, using EVs is advantageous, as vesicles protect antigens against degradation, enabling long‐term storage, and can work as potent adjuvants, eliciting immunological responses without producing a generalised reaction with substantial side effects (Tan et al. 2018; Zhu et al. 2021). We recently demonstrated that EVs from the cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis) can be safely used as nanocarriers of heterologous proteins in fish (Matinha‐Cardoso et al. 2022). Synechocystis EVs were biocompatible with zebrafish larvae, not triggering high inflammatory responses.

In this work, the stability of Synechocystis EVs and its customised protein cargo (sfGFP) were evaluated when stored under different temperature conditions and after freeze‐drying. Extracellular vesicle morphology, amount, total protein, and LPS content were investigated, as well as the fluorescent activity of the reporter protein. Moreover, the possibility of using Synechocystis EVs as a biotechnology tool to stimulate specific and adaptive immune responses in fish was assessed by following the total and specific immunoglobulins produced in European seabass. Our work highlights subunit fish vaccination and cyanobacterial EVs as a valid alternative for antigen administration.

2. Materials and Methods

2.1. Cyanobacterial Cultivation

The cyanobacterial strain Synechocystis sp. PCC 6803 EV‐trc‐GFP (Lima et al. 2022; Matinha‐Cardoso et al. 2022) was routinely cultivated in BG‐11 medium (Rippka et al. 1979) supplemented with kanamycin (100 μg mL⁻¹) in an orbital shaker (90 rpm) at 26°C, under a 16 h light (20 μmol photons m⁻² s⁻¹)/8 h dark cycle regimen. Growth was monitored spectrophotometrically by following optical density at 730 nm (OD₇₃₀).

2.2. Extracellular Vesicles Analysis

To assess EV morphology, size (in nm), and amount (reported here both as number of particles and as concentration; the latter is expressed as number of particles mL⁻¹ OD₇₃₀ ⁻¹, due to the need of normalising the amount of EVs among samples obtained in different cultivations, with OD₇₃₀ used as a proxy to evaluate amount of cells), isolated EVs were analysed by different techniques, including nanoparticle tracking analysis (NTA) (NanoSight), transmission electron microscopy (TEM) of negatively stained samples, and lipopolysaccharide profile analysis by electrophoretic separation in SDS‐polyacrylamide gels, as detailed in Biller et al. (2022). Moreover, quantification of proteins in EV samples was performed by an SDS‐polyacrylamide gel electrophoresis‐based method, using bovine serum albumin (BSA) as standard, and stained with colloidal Coomassie Brilliant Blue (Sigma). BSA samples (containing 0, 0.125, 0.5, 1, 2, and 5 μg) were prepared and loaded side by side with isolated EVs samples. A GS‐800 Calibrated Densitometer (BioRad) was used to scan the Coomassie blue‐stained gels, and densitometry analysis was carried out to determine protein concentration in EVs samples (μg μL⁻¹) using ImageJ software (https://imagej.net/ij/).

2.3. Evaluation of EV and EV‐Cargo Stability Under Different Storage Conditions and After Freeze‐Drying

To evaluate the stability of EV‐trc‐GFP extracellular vesicles and packaged heterologous cargo when stored under different conditions, strain EV‐trc‐GFP was initially cultivated in BG‐11 (without antibiotic supplementation) and sparged with air (1 L min⁻¹), in 500 mL glass gas washing bottles, in a final volume of 200 mL of culture. Cultivation was carried out at 28°C under an irradiance of 40–50 μmol photons m⁻² s⁻¹ and a 16 h light/8 h dark diel cycle until cell density reached an OD₇₃₀ of approximately 1.0–1.2. Cultures were then reinoculated in 1 L glass gas washing bottles containing 600 mL of medium and sparged with air (1 L min⁻¹). The initial OD₇₃₀ was set to 0.05, and cultivation proceeded under the same conditions described for inoculum preparation, resulting in an OD₇₃₀ of 3.0. Then, the biomass was separated from the extracellular medium by centrifugation (4000 g, 10 min), followed by filtration through 0.45 μm pore size filters. For EV isolation, a protocol described by Biller et al. (2022) was followed. In brief, the filtered extracellular medium was concentrated by ultrafiltration using centrifugal filters with a nominal molecular weight limit of 100 kDa (Pall) to a volume of approximately 30–60 mL. Vesicles were separated from the remaining medium by ultracentrifugation (100,000 g, for 3 h, at 4°C), and the EV pellet was suspended in phosphate‐buffered saline solution (PBS). The isolated EV sample was separated in aliquots and stored immediately at either −80°C, −20°C, +4°C, or room temperature. Frozen samples were thawed on ice after 4 and 8 months under such storage conditions. Then, EV morphology was observed by transmission electron microscopy (TEM), EV size and amount evaluated by nanoparticle tracking analysis (NTA), and total protein and LPS content determined by densitometry analysis of Coomassie blue‐ or LPS‐stained SDS‐polyacrylamide gels, respectively, as detailed by Matinha‐Cardoso et al. (2022). To evaluate the level of sfGFP in EV samples stored under different conditions, a fluorometric assay was carried out as follows: EVs were loaded onto dark 96‐well microtiter plates with a clear bottom (Nunc), and sfGFP fluorescence detected on a Synergy Mx fluorimeter (BioTek) at 528 nm (10 nm bandpass) after excitation at 488 nm (10 nm bandpass). Fluorescence from the various EV samples was recorded and compared to freshly isolated sfGFP‐containing EVs, normalised by OD₇₃₀, and the total amount of vesicles. In addition to fluorescence measurements, a western blot analysis was performed on the samples after the 8‐month storage period to compare their sfGFP protein level. For that, EV samples were separated side by side on 16% (w/v) SDS‐polyacrylamide gels, followed by transfer to a PVDF membrane. A GFP‐specific monoclonal mouse antibody (Roche) was used as primary antibody (1:4000), and goat‐anti‐mouse IgG (Invitrogen) linked to horseradish peroxidase (HRP) was used as secondary antibody (1:5000). Immunodetection was performed on a ChemiDoc MP System (BioRad), using the Clarity Western ECL Substrate (BioRad) according to the instructions of the manufacturer. In addition to testing different temperature‐storing conditions, EVs and their cargo stability were evaluated after the freeze‐drying process using protein, LPS, NTA, TEM, and fluorescence methods, as described above. All these experiments were carried out with EVs isolated from three independent cultivation experiments.

2.4. Isolation of Extracellular Vesicles and Soluble sfGFP From the Extracellular Medium of EV‐trc‐GFP for Fish Experiments

As detailed earlier (Lima et al. 2022), cyanobacterial strain EV‐trc‐GFP releases sfGFP packaged in EVs but also in its soluble, free form. Therefore, both bioproducts were isolated from the extracellular medium of this strain for immunisation trials with European seabass. Growth of the strain EV‐trc‐GFP was conducted as described above by adopting batch mode cultivation. For each cultivation, 3.6 L of culture was divided into 1 L glass gas washing bottles containing 600 mL of medium. When reaching an OD₇₃₀ of 3.0, cultures were processed as described above for EV isolation, with the final pellet resuspended in sterile BG‐11. To obtain enough vesicles for downstream characterisation and European seabass immunisation, 10 separate cultivation‐ and EV isolation‐runs were performed sequentially over the period of 5 months (corresponding to a total culture volume of approximately 36 L), which were pooled in the same vial. Then, for BG‐11 salt removal, the isolated EVs were subjected to dialysis (membrane with a molecular weight cutoff of 12 kDa) against PBS solution and stored at −80°C until further analysis.

After removing the EVs from the extracellular medium obtained from the cultivation of EV‐trc‐GFP, soluble sfGFP present in the medium was isolated using anion exchange chromatography (HiTrap Q HP; Cytiva), followed by size exclusion chromatography. Later, the isolated protein was concentrated and subjected to buffer exchange to PBS by centrifugal ultrafiltration with filters with a nominal molecular weight limit of 3 kDa (Amicon; Millipore). Before being stored at −80°C until further analysis, the final protein sample was analysed by SDS‐polyacrylamide gel electrophoresis followed by Coomassie blue staining of the gel for assessment of its purity, and protein quantification was determined by the Bradford assay (BioRad).

2.5. sfGFP Quantitation in EV Samples

To determine the amount of sfGFP packaged in isolated EVs, the purified sfGFP protein sample and isolated EVs were separated side by side on 16% (w/v) SDS‐polyacrylamide gels, followed by western blotting analyses. A GFP‐specific monoclonal mouse antibody (Roche) was used as primary antibody (1:4000), and goat‐anti‐mouse IgG (Invitrogen) linked to horseradish peroxidase (HRP) was used as secondary antibody (1:5000). Immunodetection was performed on a ChemiDoc MP System (BioRad), using the Clarity Western ECL Substrate (BioRad) reagent according to the instructions of the manufacturer. Band intensity was quantified by densitometry using the ImageJ software (Matinha‐Cardoso et al. 2022). Isolated sfGFP protein was used to establish a standard curve, against which the amount of sfGFP packaged in EVs was determined, expressed in ng of sfGFP per μg of vesicular protein.

2.6. Animals and In Vivo Experimental Conditions

Animal experiments were approved by the Animal Welfare Committee of the Interdisciplinary Centre of Marine and Environmental Research (CIIMAR), carried out in a registered installation (N16091.UDER), and performed by trained scientists (following FELASA category C recommendations) in full compliance with the European Directive 2010/63/EU on the protection of animals used for scientific purposes. European seabass ( Dicentrarchus labrax ) juveniles were obtained from a commercial fish farm (Maresa S.A., Ayamonte, Huelva, Spain) and submitted to a quarantine period of 15 days before being transferred to a recirculating water system (RAS) equipped with 15 cylindrical fibreglass tanks of 100 L water capacity and thermo‐regulated to 22.0°C ± 1.0°C. Tanks were supplied with a continuous flow of filtered seawater (Salinity: 34.0 ± 1.0 g L⁻¹; Oxygen: 7.5 mg L⁻¹; ${\mathrm{NH}}_4^{+}$ ≤ 0.05 mg L⁻¹; ${\mathrm{NO}}_2^{-}$ ≤ 0.5 mg L⁻¹).

2.7. European Seabass Immunisation With sfGFP‐Loaded EVs

After 15 days of adaptation to the experimental conditions, 24 European seabass with an initial mean body weight of 26.5 g were randomly distributed to each tank and to the experimental tests that were tested in duplicate as follows: CTR (non‐vaccinated group): fish injected with 100 μL PBS; GFP: fish injected once with 16.5 μg purified sfGFP (in 100 μL); High‐dose EVs: fish injected once with isolated EVs carrying a total of 16.5 μg of sfGFP, which corresponds to 825 μg of vesicular protein; 1× Low‐dose EVs: fish injected once with isolated EVs carrying a total of 8.25 μg of sfGFP, corresponding to 412.5 μg of vesicular protein; 2× Low‐dose EVs: fish injected twice (at day 1 and 15, as boost to first injection) with 8.25 μg of sfGFP, corresponding to a total of 16.5 μg of sfGFP and 825 μg of vesicular protein. Fish were fed daily by hand, 7 days a week, until apparent visual satiation, during 30 days. Blood of six fish/tank (n = 12 fish per treatment, i.e., 12 biological replicates) was collected at days 15 and 30 post‐immunisation for serum antibody analysis.

2.8. Dot Blot Analysis and ELISA of European Seabass Serum

For determining total and sfGFP‐specific antibody production by European seabass upon injection, dot blot and ELISA analyses were carried out overnight at 4°C with 50 ng of purified GFP protein against 10 μL of immobilised undiluted fish serum (pooled from six fish) of either immunised (fish injected with GFP, or GFP‐carrying EVs) or non‐immunised (control fish injected with PBS only) animals, or bovine serum albumin (BSA) as control for unspecific binding. An anti‐GFP specific primary antibody (Roche) was used, followed by incubation with goat anti‐rabbit IgG HRP‐conjugated secondary antibody (Invitrogen). Washing with PBS‐0.1% (v/v) Tween was included between each step. Blots were visualised in a ChemiDoc XRS Gel Imaging System (Bio‐Rad) and analysed with the Image Lab Software (Bio‐Rad). For ELISA, 96‐well microtiter plates (Nunc Flat bottom MaxiSorp) were coated with 100 μL of a 20 μg μL⁻¹ purified GFP solution in PBS and incubated at room temperature for 4 h (uncoated wells were used as blank, and wells coated with 20 μg μL⁻¹ BSA solution were also included). Following a 2 h blocking step with 200 μL of PBS‐10% (w/v) low‐fat milk, wells were coated with 100 μL of fish serum diluted 1:6 in PBS (n = 12) and incubated overnight at 4°C. On the following day, wells were incubated for 1 h with 100 μL of anti‐European seabass IgM monoclonal antibody (Aquatic Diagnostics) diluted 1:33 in PBS, followed by 1 h with a goat anti‐Mouse IgG (H + L) HRP‐conjugated secondary antibody (Invitrogen) diluted 1:10,000 in PBS. Wells were washed three times with PBS‐0.1% (v/v) Tween between each step. After washing, 100 μL of TMB (3,3′,5,5′‐tetramethylbenzidine) were added and incubated for 20 min. The reaction was stopped by adding 100 μL of 2 M H₂SO₄, and absorbance was measured immediately at 450 nm. For total antibody measurement, 96‐well microtiter plates were coated with 100 μL of fish serum solution diluted in PBS (1:7500) and incubated overnight at 4°C (uncoated wells were used as blank controls, and wells coated with 20 μg μL⁻¹ BSA solution were also included). After that, blocking and incubation with anti‐European seabass IgM monoclonal antibody and secondary antibody, as well as respective washing steps and signal detection, were performed as described above. ELISA assays were repeated three times (i.e., three technical replicates).

2.9. Data Analysis

Statistical analysis was done using the SPSS 26.0 software package for Windows or GraphPad Prism 8. Data was tested for normality using Shapiro–Wilk normality tests before determining group differences using non‐parametric Kruskal–Wallis or one‐way ANOVA analysis, depending on the data normality. Tukey's post hoc test was used to determine significant differences among groups in fish experiments. A p‐value of ≤ 0.05 was considered indicative of statistical significance in all analyses.

3. Results

3.1. Synechocystis EVs Are Stable Under Different Storage Conditions, Retaining sfGFP‐Protein Fluorescence Activity

To understand the possible effects on physicochemical properties of cyanobacterial EVs caused by storing them under different conditions, EVs from the cyanobacterium Synechocystis strain EV‐trc‐GFP were stored under freezing temperatures (−80°C and −20°C), at 4°C, or room temperature (RT). After 4 and 8 months under such conditions, EVs were thoroughly characterised to evaluate their biophysical (morphology, amount, and size) and biochemical features (lipopolysaccharide [LPS] and protein profiles), and the data were compared to that of freshly prepared EV samples isolated from the same strain. In addition, their heterologous protein cargo (sfGFP) stability was also assessed through fluorescence measurements and western blot analysis. The stability of EVs and their cargo after freeze‐drying conditions was also investigated.

Transmission electron microscopy analysis of negatively stained EVs showed the typical spherical shape of vesicles in all samples (Figure 1A), with no discernable differences in size and morphology. Still, vesicles stored at RT showed some degree of aggregation.

FIGURE 1.

Biophysical parameters of vesicles isolated from Synechocystis EV‐trc‐GFP strain upon lyophilization and storage under different temperature conditions. Assessment of the stability of EVs from Synechocystis EV‐trc‐GFP when subjected to freeze‐drying (FD) or different temperature storing conditions (−80°C, −20°C, +4°C, and room temperature [RT]) after 4 and 8 months. (A) Negative staining, transmission electron microscopy (scale bars—200 nm); (B) number of particles normalised by the volume and OD₇₃₀ of cultures (upper panel) and particle mean size (lower panel) determined by NTA. T0—freshly isolated vesicles. Error bars represent the standard deviations of three independent biological replicates. **p < 0.01.

Nanoparticle tracking analysis indicated no differences in particle concentration in the different treatments, with the number of particles mL⁻¹ OD₇₃₀ ⁻¹ and their mean size within the range of freshly isolated vesicles (2.0E⁺⁰⁹ ± 3.2E⁺⁰⁸ particles mL⁻¹ OD₇₃₀ ⁻¹ and 168 ± 10 nm, respectively) (Figure 1B). Only for EVs stored at RT for 8 months could a small increase in size be detected (Figure 1B).

Regarding LPS and protein analysis through SDS‐polyacrylamide gels, no signs of degradation of these biomolecules were detected upon storage, as indicated by the sharp aspect of the bands (Figures S1 and S2), and quantification, with no apparent loss of material (Figure 2A,B). Concerning cargo stability, the reporter sfGFP maintained its integrity and fluorescence capacity when vesicles were stored at –80°C and –20°C, as indicated by western blot analysis (Figure S2), and fluorescence measurements after 4‐ and 8‐month storage periods (Figure 2C). As for EVs stored at 4°C and RT, no statistical differences were observed compared to the freshly isolated EVs. However, these samples presented increased variability compared to the other groups (Figure 2C; Figure S2). In all the tested stability metrics, no differences between freeze‐dried (FD) samples were found compared to the fresh samples (Figures 1 and 2).

FIGURE 2.

Analysis of biochemical parameters of vesicles isolated from the Synechocystis EV‐trc‐GFP strain upon lyophilization and storage under different temperature conditions. Assessment of the stability of EVs from Synechocystis EV‐trc‐GFP when subjected to freeze‐drying (FD) or different temperature storing conditions (−80°C, −20°C, +4°C and room temperature [RT]) after 4 and 8 months. (A) Protein quantification normalised by the volume and OD₇₃₀ of cultures; (B) lipopolysaccharides quantification normalised by the volume and OD₇₃₀ of cultures; (C) fluorescence measurements normalised by the number of particles; T0—freshly isolated vesicles. Error bars represent the standard deviations of three independent biological replicates.

3.2. Preparation of Soluble sfGFP and sfGFP‐Loaded EVs for European Seabass Immunisation

To evaluate the capacity of cyanobacterial‐derived EVs to work as antigen delivery vehicles for European seabass immunisation and consequent antibody production, sfGFP‐loaded vesicles from the Synechocystis strain EV‐trc‐GFP were used. The aspect of the EV pellet after ultracentrifugation can be observed in Figure 3A. Vesicles from the different batches, each stored at −80°C after the isolation step, were pooled in a single vial and suspended in sterile PBS solution, resulting in approximately 12.5 mL of sfGFP‐loaded EV suspension, which upon inspection by TEM showed the presence of numerous EVs (Figure 3A).

FIGURE 3.

Isolation and characterisation of Synechocystis EV‐trc‐GFP vesicles for immunisation trials in European seabass juveniles. (A) Representative image of a typical EV pellet (panel to the left) after ultracentrifugation; EV suspension from a pool of 36 L of cyanobacterial culture (panel in the centre), with the respective transmission electron microscopy analysis (panel to the right); (B) purified sfGFP from Synechocystis EV‐trc‐GFP extracellular medium (purified by anion exchange chromatography, size exclusion chromatography and ultrafiltration), and (C) respective protein profile in a Coomassie‐stained SDS‐polyacrylamide gel. (D) Quantification of sfGFP in Synechocystis EV‐trc‐GFP EVs through western blot analysis (top panel) with the respective Coomassie‐stained SDS‐polyacrylamide gel (bottom panel).

Synechocystis strain EV‐trc‐GFP also releases sfGFP freely to the medium, representing up to 90% of all the sfGFP released to the extracellular space (Lima et al. 2022). Therefore, taking advantage of the low protein complexity of the extracellular medium upon cyanobacterial cultivation and EV removal, the EV‐trc‐GFP cultivation medium was further processed to purify soluble sfGFP. The objective was to get hold of a protein that could be used as a control for European seabass immunisation (see below) and as a standard for quantification of the sfGFP content in isolated vesicles. Therefore, approximately 10 L of EV‐free extracellular medium was concentrated by centrifugal ultrafiltration through filters with a nominal molecular weight limit of 3 kDa. The concentrate was then subjected to anion exchange and size exclusion chromatography, resulting in a highly pure preparation of soluble sfGFP protein (Figure 3B,C). A western blot was then performed to quantify the amount of sfGFP protein loaded into Synechocystis EVs, using the sfGFP purified sample to generate a standard protein curve (Figure 3D). Densitometry analysis revealed that the isolated EV sample contained approximately 20 ng of sfGFP per μg of vesicular protein (20.15 ± 5.69 ng sfGFP μg⁻¹ of vesicular protein; five replicates).

3.3. European Seabass Produces Total and sfGFP‐Specific Antibodies in Response to sfGFP‐Loaded EVs but Not to Isolated sfGFP Protein

Purified sfGFP protein (16.5 μg) and different amounts of isolated Synechocystis EVs carrying sfGFP were used to inject European seabass juveniles intraperitoneally (i.p.). To evaluate dosage‐dependent responses, the immunisation scheme (Figure 4) included injection of two EV amounts: a high dose of EVs, corresponding to 825 μg of vesicular protein that carried 16.5 μg of sfGFP, and a low dose of EVs, corresponding to 412.5 μg of vesicular protein that carried 8.25 μg of sfGFP. The low dose of EVs was also given as a boost 15 days after the first injection to assess the advantages of two versus single immunisations. Fifteen and 30 days after injection, both immunised and non‐immunised fish were sacrificed, and their sera were used to determine total and specific anti‐GFP antibodies production.

FIGURE 4.

Schematic representation of the European seabass immunisation strategy to evaluate the immunomodulatory capacity of cyanobacterial extracellular vesicles as antigen delivery vehicles. Purified sfGFP protein (16.5 μg) and different amounts of isolated Synechocystis EV‐trc‐GFP vesicles carrying sfGFP were used for intraperitoneal injection of European seabass juveniles. The immunisation scheme included injection of two EV amounts: a high‐dose, corresponding to 825 μg of vesicular protein that carried 16.5 μg of sfGFP, and a low‐dose of EVs, corresponding to 412.5 μg of vesicular protein that carried 8.25 μg of sfGFP. Low‐dose EVs were also given 15 days after the first injection as a boost to assess the advantages of two versus single immunisations. Immunised and non‐immunised (CTR, injection with phosphate‐buffered saline solution [PBS]) fish were sacrificed 15 and 30 days after immunisation, and sera were used to determine total IgM and anti‐GFP antibody production.

Dot‐blot analysis using immobilised fish serum (from a pool of sera collected on day 30) showed an EV‐concentration‐dependent signal in immunised fish (Figure S3). While no signal was detected for non‐immunised fish, a clear signal was obtained in the sera of fish immunised both with the lowest and the highest amount of EVs (the latter showed a signal approximately 50%–60% stronger than the former, as determined by densitometry). Individual fish responses were further analysed by ELISA (Figure 5). No differences were observed between treatments in serum levels of anti‐sfGFP antibodies at day 15 post‐immunisation. However, at day 30, there was a significantly higher titre of anti‐sfGFP antibodies in the serum of fish immunised with high‐dose EVs and 2× low‐dose EVs treatment (fish immunised twice with the low‐dose) compared to non‐immunised fish (determined by one‐way ANOVA, followed by Tukey's post hoc test; p‐value was < 0.0001). In contrast, immunisation of fish with free‐sfGFP did not induce any alteration in serum levels of anti‐GFP antibodies. Fish immunised once with low‐dose EVs (1× low‐dose EVs treatment) showed intermediate antibody levels (Figure 5A). As for total IgM levels, the same tendency was observed after ELISA (Figure 5B): sfGFP alone did not alter total antibody production compared to the control, while EVs carrying sfGFP significantly increased IgM production. In this case, on day 15, the difference was already notorious for the higher dose treatment. After 30 days post‐injection, the lower EV dose with the double‐injection treatment strategy resulted in similar antibody levels compared to the higher dose condition, which remained significantly higher than when administering sfGFP alone or compared to the control (differences were statistically tested using one‐way ANOVA and Tukey's post hoc test, p‐value < 0.0001). Compared to the specific antibody‐production response, the low‐dose EVs with the single injection strategy also resulted in an intermediate level of total IgM (Figure 5B).

FIGURE 5.

sfGFP loaded in cyanobacterial extracellular vesicles triggers specific and total antibody production in European seabass juveniles upon intraperitoneal injection. Individual fish anti‐GFP [(A) GFP‐specific antibodies] and total antibody [(B) total antibodies] production after injection of the control solution (PBS), free‐sfGFP, and EVs carrying sfGFP administered in high and low doses (with 1× or 2× treatment strategies for the low dose). Antibody levels were analysed by ELISA at days 15 (T15) and 30 (T30) post‐immunisation. Differences between groups were analysed by one‐way ANOVA, followed by Tukey post hoc test. Different lowercase letters indicate statistically significant differences between groups (p < 0.05).

4. Discussion

Fish aquaculture is currently the fastest‐growing food‐producing sector, accounting for approximately 50% of the global fish consumption (FAO 2020). However, disease outbreaks caused by pathogenic microorganisms generate considerable economic losses, representing a significant limiting factor for further fish farming development. Aiming to reduce antibiotic use and mitigate antimicrobial resistance, this work unveils the potential of extracellular vesicles from the model cyanobacterium Synechocystis sp. PCC 6803, as an antigen‐carrying system, towards developing an alternative vaccination platform for aquaculture fish. To this end, EVs loaded with the reporter green fluorescent protein (sfGFP) isolated from the extracellular medium of engineered Synechocystis strain EV‐trc‐GFP (Matinha‐Cardoso et al. 2022) were assessed for stability under different storage conditions and after freeze‐drying. Furthermore, they were evaluated regarding their ability to induce sfGFP‐specific antibody and total IgM production upon injection in European seabass.

An often overlooked but critical aspect of vaccine development relates to its storage stability. The same applies to EVs, regardless of their application. Despite the growing interest in EVs, studies on their long‐term storage stability remain elusive and disputable, with most existing data focusing on mammalian‐derived vesicles (Jeyaram and Jay 2017; Kusuma et al. 2018; Richter et al. 2019; Yuan et al. 2021; Görgens et al. 2022; Sivanantham and Jin 2022; Ahmadian et al. 2024), which are generally more sensitive than bacterial EVs (Jeyaram and Jay 2017; Schulz et al. 2020). This difference likely stems from the greater resilience bacteria require to endure challenging and quickly changing environmental conditions. Here, EVs derived from the cyanobacterial strain Synechocystis EV‐trc‐GFP demonstrated remarkable stability under various storage conditions (−80°C, −20°C, 4°C, and room temperature) over 4 and 8 months. Among the stability metrics assessed here (particle size and concentration, protein content, and lipopolysaccharide levels), a significant increase in vesicle size was observed only at room temperature, likely due to particle aggregation as evidenced by transmission electron microscopy.

Furthermore, fluorescent measurements suggest that EVs carrying sfGFP can retain heterologous protein functionality despite the storage temperature, which is in agreement with previous studies reporting the cargo protective capacity of EVs (Bonnington and Kuehn 2014; Alves et al. 2016; Frank et al. 2018). Although still disputed, it is generally assumed that long‐term EVs storage at −80°C is the best condition (Jeyaram and Jay 2017), with the other options (−20°C, 4°C, and RT) affecting vesicle and cargo integrity, depending on their source. Nevertheless, Synechocystis EVs were stable even when stored at temperatures higher than −80°C, an important consideration for logistics and cost‐effective vaccine production. As an alternative to storage in liquid form, freeze‐drying (lyophilization)—a widely adopted method for stabilising thermolabile materials—offers the possibility to improve EV shelf‐life by removing water and reducing dependence on the cold chain (Guarro et al. 2022). However, this process carries the risk of compromising EV structure and cargo, with most studies reporting the need for stabilisers (e.g., glucose, lactose, sucrose, or trehalose) to preserve vesicle integrity (Kusuma et al. 2018; Yuan et al. 2021; Guarro et al. 2022; Trenkenschuh et al. 2022; Susa et al. 2023; Adamo et al. 2025). In this work, however, EV‐trc‐GFP strain vesicles retained their structural integrity and cargo stability post‐lyophilization without using any stabilisers. This highlights the intrinsic robustness of cyanobacterial EVs and reinforces their potential as a resilient and practical platform for antigen delivery and, thus, vaccine development.

Although bacterial EVs have been extensively studied as vaccination platforms for humans and terrestrial animals, their application in aquaculture remains underexplored. Only a few studies have assessed their potential in fish vaccination and protection, and these have exclusively employed EVs isolated directly from pathogenic bacteria (Park et al. 2011; Brudal et al. 2015; Tandberg et al. 2017; Maiti et al. 2020; Erfanmanesh et al. 2022; Teixeira et al. 2023; Escribano et al. 2024) instead of using engineered EVs derived from non‐pathogenic species. Developing vaccines against all pathogenic bacteria with homologous EVs is conceptually possible, as all bacteria are thought to release EVs (Coelho and Casadevall 2019). However, in addition to safety and handling issues, not all pathogenic species are equally suitable for EV production and isolation. Furthermore, due to the stable virulent factors present in bacterial EVs, they are frequently associated with toxicity to the host (Park et al. 2010; Qing et al. 2019; Lima et al. 2020). In a new era with ever‐increasing synthetic biology tools, researchers focus on engineering non‐pathogenic species to produce safe EVs with tailored lumen and/or surface content (Kim et al. 2008; Gerritzen et al. 2017; Qing et al. 2019; Cheng et al. 2021; Ren et al. 2021; Richter et al. 2021). This way, vesicles from high‐producing non‐pathogenic species can be developed as universal platforms to deliver proteins of interest for various applications, including fish vaccination.

Cyanobacteria are non‐pathogenic prokaryotes considered one of the greenest and most sustainable microbial cell factories of relevant bio‐products (Ruffing 2011; Wijffels et al. 2013; Zahra et al. 2020; Khalifa et al. 2021; Jester et al. 2022). Recently, we engineered the model cyanobacterium Synechocystis sp. PCC 6803 to generate EV‐trc‐GFP, a hyper‐vesiculating strain capable of producing fish‐biocompatible EVs carrying the reporter sfGFP (Matinha‐Cardoso et al. 2022). The present work provides compelling evidence that Synechocystis EVs can be used as effective antigen delivery systems in fish through injection.

EVs carrying sfGFP from the EV‐trc‐GFP strain elicited a robust adaptive immune response, significantly increasing the production of sfGFP‐specific immunoglobulins compared to fish injected with the control PBS solution. The same happened when comparing with the reporter protein in purified, soluble form, indicating that EVs may not only serve as antigen carriers but also function as adjuvants that enhance the immunogenicity of their cargo. Total IgM measurements exhibited the same trend, with elevated antibody levels detected as early as 15 days post‐injection in fish immunised with EVs compared to the control and soluble sfGFP treatments. This highlights the adjuvant capacity of these vesicles, in agreement with previous reports that document the immunostimulatory capacity of bacterial EVs, not only in fish (Hong et al. 2009; Chapagain et al. 2024; Oliver et al. 2023; Dias et al. 2024; Jayathilaka et al. 2024; Vicente‐Gil et al. 2024; Dias et al. 2025) but across other animal models and even in humans (Tartaglia et al. 2018; Molina‐Tijeras et al. 2019; Morishita et al. 2023; Sharifpour et al. 2024). In this regard, it is noteworthy that EVs from the most extensively cultivated cyanobacterium in large‐scale production, Spirulina (a widely accepted common name for species from the genus Limnospira), have recently been reported in mice to present remarkable properties as an adjuvant for subunit vaccines (Sharifpour et al. 2024). Adjuvants are indispensable components of vaccines, but their mode of action is not fully clear yet (Zhao et al. 2023); nevertheless, it is accepted that adjuvants promote the generation of antigen‐presentation and co‐stimulatory signals by activating antigen‐presenting cells (Zhao et al. 2023; Sharifpour et al. 2024). Because bacterial EVs retain the physiochemical characteristics of the bacteria from which they are derived, the presence of microbial‐associated molecular patterns (MAMPs) is described to initiate and facilitate the interaction process with the pattern recognition receptors (PRRs) expressed by the immune cells (Zhu et al. 2021). This, in turn, facilitates antigen uptake and initiates signal cascades that lead to the host immune response, which, if not excessively activated, can stimulate safe and effective adaptive immunity. This may explain the higher efficiency of Synechocystis vesicles carrying the sfGFP in eliciting immune responses compared to the other treatments tested in this work. It should be noted that although GFP is generally recognised as a low immunogenic protein (Skelton et al. 2001; Ansari et al. 2016), it has already been shown to increase its immunogenicity in mice when delivered in EVs (Chen et al. 2010; Hu et al. 2022). The protein sfGFP was used here as a model antigen, and while weakly immunogenic, its use was justified as a proof‐of‐concept; whether the effect played by EVs is similar when using pathogen‐derived antigens remains to be determined. In addition to antibody production, additional studies should equally address the expression of relevant immune molecular markers to better understand the mechanisms of action.

When comparing the different EV administration strategies employed here, a single high‐dose EV injection was as effective in stimulating specific and total antibody production as a two‐dose treatment with a lower EV concentration, administered on day 1 and again on day 15. Therefore, the antigen dose, rather than the number of administrations, seemed to be the key factor in maximising the immune response, at least against sfGFP, a finding with significant implications for the cost, labor, and animal handling associated with large‐scale fish vaccination. Additionally, the administration of an EV load as high as 825 μg of vesicular protein, resulting in no apparent impact on fish fitness (assessed by the lack of fish mortality and animal behaviour change), further highlights the high degree of cyanobacterial EV biocompatibility in fish (Matinha‐Cardoso et al. 2022). This is particularly relevant for engineered vesicles with tailored content since it corresponds to a 33‐fold higher EV dosage than that used for EVs of an isolated pathogen (Teixeira et al. 2023). This feature allows for exploring various immunisation dosages with cyanobacterial EVs for effective protection.

Another aspect that warrants further investigation is the amount of heterologous protein that can be loaded in cyanobacterial EVs: here, we report that only a modest amount of sfGFP was loaded (approximately 20 ng per μg of vesicular protein). Various approaches could be considered to increase the overall amount of target protein in EVs, some of which have already been experimentally tested in other systems. On one hand, one could consider using a promoter of higher strength. This could result in the accumulation of more protein in the periplasm, creating an increase in periplasmic pressure (Schwechheimer et al. 2014), forcing its entry into EVs. However, this may represent a burden to the cell as it overloads the protein translocation system to the periplasm, impacting transport of native proteins (some involved in fundamental biological processes, such as nutrient uptake). On the other hand, one could also consider using a split protein bioconjugation system to create a synthetic linkage between the protein of interest and an abundant membrane protein to facilitate loading of the EV (as described, e.g., by Alves et al. 2016). Regardless of the approach, future studies should clarify the amount of target protein in EV to produce the most effective immune response in the host.

This study sets the ground for applying cyanobacterial EVs as a safe and stable antigen carrier platform for fish vaccination. Direct protection against infection could not be assessed here due to using sfGFP as a reporter immunogen rather than a pathogenic antigen. However, future studies should assess immune protection after challenges in different fish species when loading EVs with immunogenic antigens from relevant pathogenic bacteria affecting aquaculture facilities.

Taking advantage of the demonstrated resilience of cyanobacterial EVs under different temperature storing conditions and after freeze‐drying, and considering the industry's demand for oral solutions, the effect of different administration routes (oral vs. injection) should also be compared and investigated.

Author Contributions

Jorge Matinha‐Cardoso: conceptualization, methodology, investigation, data curation, writing – review and editing, writing – original draft. Gabriela Gonçalves: conceptualization, methodology, investigation, writing – review and editing. Filipe Coutinho: conceptualization, methodology, investigation, writing – review and editing. Steeve Lima: conceptualization, methodology, investigation, writing – review and editing. Lourenço Bonneville: methodology, investigation. Mónica Serrano: supervision, funding acquisition, writing – review and editing. Paula Tamagnini: supervision, funding acquisition, resources, writing – review and editing. Aires Oliva‐Teles: supervision, funding acquisition, resources, writing – review and editing. Ana Couto: methodology, investigation, supervision, writing – review and editing. Cláudia R. Serra: conceptualization, methodology, investigation, data curation, supervision, funding acquisition, project administration, writing – review and editing. Paulo Oliveira: conceptualization, methodology, investigation, data curation, supervision, funding acquisition, project administration, writing – original draft, writing – review and editing.

Funding

This work was supported by national funds through Fundação para a Ciência e a Tecnologia, I.P. (FCT) and by the European Commission's Recovery and Resilience Facility, within the scope of UIDB/04423/2025, UID/PRR/04423/2025 (https://doi.org/10.54499/UID/PRR/04423/2025) and LA/P/0101/2020 (https://doi.org/10.54499/LA/P/0101/2020); European Regional Development Fund (ERDF) through the COMPETE 2020 Operacional Programme for Competitiveness and Internationalisation (POCI), Portugal 2020, and by Portuguese funds through FCT in the framework of projects POCI‐01‐0145‐FEDER‐029540 (PTDC/BIA‐OUT/29540/2017) and PTDC/CVT‐CVT/2477/2021; ERDF through the Innovation and Digital Transition Programme (COMPETE 2030) and by national funds through FCT within the scope of project COMPETE2030‐FEDER‐00842400 (No. 14867) (https://doi.org/10.54499/2023.17583.ICDT); and by the operation PremiumAlgae (02185500), supported by COMPETE 2030, Algarve 2030, Portugal 2030 and by the European Union. Additional financial support was obtained from the European Union's Horizon 2020 Research and Innovation programme under grant agreement no. 952374 (BB4F), and NORTE2030‐FEDER‐01796500, project co‐funded by the European Union through the NORTE 2030 Regional Program (BB4F‐Complementar). FCT also granted PhD fellowships: SFRH/BD/130478/2017 (S.L.), 2021.07724.BD (G.G.), and 2022.11873.BD (J.M.‐C).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: mbt270294‐sup‐0001‐DataS1.docx.

Matinha‐Cardoso, J. , Gonçalves G., Coutinho F., et al. 2026. “Cyanobacterial Extracellular Vesicles as Protein Carriers: Towards Fish Vaccination.” Microbial Biotechnology 19, no. 1: e70294. 10.1111/1751-7915.70294.

Data Availability Statement

The data that supports the findings of this study is available in the Supporting Information of this article.