Extracellular vesicles induce protective immunity against Trichuris muris

Abstract Gastrointestinal nematodes, such as Trichuris trichiura (human whipworm), are a major source of morbidity in humans and their livestock. There is a paucity of commercially available vaccines against these parasites, and vaccine development for T. trichiura has been impeded by a lack of known host protective antigens. Experimental vaccinations with T. muris (murine whipworm) soluble Excretory/Secretory (ES) material have demonstrated that it is possible to induce protective immunity in mice; however, the potential for extracellular vesicles (EVs) as a source of antigenic material has remained relatively unexplored. Here, we demonstrate that EVs isolated from T. muris ES can induce protective immunity in mice when administered as a vaccine without adjuvant and show that the protective properties of these EVs are dependent on intact vesicles. We also identified several proteins within EV preparations that are targeted by the host antibodies following vaccination and subsequent infection with T. muris. Many of these proteins, including VWD and vitellogenin N and DUF1943‐domain‐containing protein, vacuolar protein sorting‐associated protein 52 and TSP‐1 domain‐containing protein, were detected in both soluble ES and EV samples and have homologues in other parasites of medical and veterinary importance, and as such are possible protective antigens.

mice from a subsequent T. muris infection when administered as a vaccine without adjuvant. This suggests that Trichuris EVs are a viable source of host protective components and that administration of recombinant Trichuris antigens within EVs may be an effective alternative to traditional vaccines formulated with adjuvant. These studies using T. muris will help inform vaccine design for T. trichiura.

| Maintenance of animals and parasites
C57BL/6 (Envigo, UK) and SCID (University of Manchester) mice were maintained in individually ventilated cages at 22 ± 1°C and 65% humidity with a 12 hour light-dark cycle. Mice had free access to food and water, and all procedures were carried out on mice 6-8 weeks of age or older, under the Home Office Scientific Procedures Act (1986). All experiments were carried out under project licence 70/8127 and conformed with the University of Manchester Animal Welfare and Ethical Review Body (AWERB) and ARRIVE guidelines. Animals were humanely killed by CO 2 inhalation followed by terminal exsanguination or cervical dislocation. The Edinburgh (E) strain of T. muris was used for all experiments and parasite maintenance was carried out as described previously. 24

| Isolation of EVs
ES was collected by culturing adult parasites (day 35 to 42 postinfection) in RPMI media supplemented with 500 U/mL penicillin and 500 μg/mL streptomycin (Sigma). Supernatants from worm cultures were collected after 4 and 18 hours and were centrifuged at 720 g for 15 minutes to separate the eggs (pellet) from the ES (supernatant). Supernatants were filtered using a 0.22 μm filter (Millipore) to remove cellular debris and microvesicles, and EVs were isolated by ultracentrifugation at 100 000 g for 2 hours in polyallomer tubes (Beckman Coulter). The EV pellet was washed by ultracentrifugation at 100 000 g for 2 hours in PBS. EV pellets were resuspended in 2 mL PBS and stored at −20°C until required.

| TEM analysis of EV samples
Samples were transferred to formvar-carbon-coated EM grids and negatively stained with 2% (w/v) uranyl acetate. Samples were imaged using a Tecnai BioTwin microscope, at 100 Kv under low-dose conditions. Images were recorded using a Gatan Orius CCD camera at 3.5 Å/pixel. ImageJ v1.46r (National Institute of Health) was used to view images and to add scale bars.

| Dynamic light scattering (DLS) of EVs
Dynamic light scattering was used to measure the size distributions of the EV preparations. DLS measurements were performed using a Zetasizer Nano S (Malvern) at a controlled temperature of 25°C.
Three measurements of 13 averages were taken and the number distribution of particles is reported.

| Proteomic analysis of EVs
Preparation of EV samples for tryptic digestion was carried out as described by Marcilla and colleagues 25 and mass spectrometry analysis was carried out as described previously. 26 The results were analysed using Scaffold Proteome Software (Scaffold, USA) and the exclusive unique peptide count was displayed for each protein (criteria set to 95% protein threshold, 50% peptide threshold, minimum 2 peptides identified). Proteins identified in two out of the three samples were listed. The SignalP Server version 4.1 (http://www.cbs.dtu. dk/services/SignalP/, Technical University of Denmark) was used to predict whether proteins had signal peptides. The protein content of T. muris EVs was also compared to that of T. muris ES 14 from which it was purified.

| Vaccination studies
All vaccination studies were carried out in male C57BL/6 mice. Prior to each vaccination study, the amount of protein in each EV sample was measured using a bicinchoninic acid assay kit. Mice were vaccinated subcutaneously with 3 μg of material (either lysed or whole

| EV lysis and protein quantification
EVs were lysed by adding 0.1% (v/v) SDS, followed by three freeze/ thaw cycles, whereby vesicles were frozen in liquid nitrogen and thawed in a 37°C water bath, with vigorous vortexing between each step. Lysis was confirmed using DLS, as described above. The amount of protein in each sample was measured using a bicinchoninic acid assay, according to the manufacturer's instructions. Protein concentration was used to standardize EV vaccinations.

| Characterization of EVs isolated from T. muris ES
EVs were isolated from T. muris ES by ultracentrifugation at 100 000 g for 2 hours. Pelleted material was viewed by transmission electron microscopy, and a heterogeneous population of cup-shaped vesicles, approximately 30-100 nm in diameter, was observed ( Figure 1A). DLS analysis also confirmed that the majority (96.5%) of vesicles isolated from T. muris ES were between 30 and 100 nm in diameter and this was consistent between samples ( Figure 1B). The size and shape of these EVs are typical of exosomes. 29,30 Proteomic analysis revealed the presence of 125 proteins within T. muris EV samples (Table S1). A number of known exosome markers were identified, including tetraspanins (tetraspanin 9 and TSP-1 domain-containing protein), heat shock proteins, enolase, Rab proteins, and apoptosis-linked gene 2 interacting protein X 1 (Alix, Table 1, references. 19,30 These data suggest that the vesicles isolated by ultracentrifugation of ES are likely to be exosomes. Proteomic analysis also revealed that 23% of EV proteins were not present in T. muris ES and 76% of these proteins lack a signal peptide (68% of total EV proteins, Table S1).

| Vaccination with T. muris EVs can induce protective immunity without adjuvant and protection is dependent on intact vesicles
In order to investigate whether T. muris EVs contain antigenic material capable of stimulating protective immunity, male C57BL/6 mice were subcutaneously vaccinated with 3 μg of isolated EVs, followed Scale bars denote 100 nm, and image is representative of three preparations. B, shows the size profile of a typical T. muris EV sample, as measured by DLS. EVs resulted in a statistically significant reduction in worm burden compared to the sham vaccination group (vaccinated with PBS only, P = .0001, Figure 2A). Importantly, the mean worm burden for mice vaccinated with lysed EVs was similar to that of the sham vaccination group (P = .0754, Figure 2A), demonstrating that intact vesicles are required to stimulate protective immunity.

| Vaccination with EVs boosts IgG1 serum antibody response to soluble ES components
Antiparasite IgG1 and IgG2a/c serum antibodies are often used as surrogate markers of resistance/chronicity during T. muris infection. 31 The serum IgG1 and IgG2a/c antibody response against ES depleted of EVs was measured for each vaccination group.
Significantly higher IgG1 antibody levels were measured for the EV vaccination group compared to the sham vaccination group (P = .0001, Figure 2B). High levels of antiparasite IgG2a/c, were also measured for the EV vaccination group ( Figure 2C), which may suggest that EV vaccinated mice mount a mixed Th1/Th2 response, or perhaps that the infection was expelled more slowly compared to the ES vaccination group. High levels of antiparasite IgG1 ( Figure 2B) and low levels of antiparasite IgG2a/c ( Figure 2C) were detected for the ES vaccination group, confirming that successful vaccination stimulates Th2 immunity, while high levels of antiparasite IgG2a/c antibodies were measured for the sham vaccination group, confirming that low-dose infection naturally primes for chronicity ( Figure 2C).

| Identification of EV components targeted by serum IgG antibodies following vaccination
Western blotting was performed to investigate which EV and ES components are recognized by serum IgG antibodies following vaccination of mice with PBS (sham), EVs or ES and subsequent T. muris infection ( Figure 3A-C). Infection alone does not generate IgG antibodies against EV material ( Figure 3A), however, vaccination with EVs primes for IgG antibodies that target a range of EV components between 50 and 200 kDa in size (indicated by asterisks in Figure 3B).
Sera collected from the ES vaccination group contained IgG antibodies that target 80 and 100 kDa EV components (indicated by asterisks in Figure 3C). Sera taken from all three groups also recognized a wide range of ES components ( Figure 3A-C).  Figure 3D shows SDS-PAGE separation of EV and ES material. Bands corresponding to 100, 80 and 70 kDa EV components were excised from the gel, since these were the most prominent bands in Figure 3B. The protein composition of these bands was determined by mass spectrometry (Table 2)  Bands corresponding to 100 (Band 1), 80 (Band 2) and 70 kDa (Band 3) were excised from the SDS-PAGE gel shown in Figure 3D, and their protein content was analysed by mass spectrometry. The proteins identified within these bands are listed. The number of unique peptides identified for each protein is displayed (criteria set to 95% protein threshold, 50% peptide threshold). a Indicates proteins identified within ES depleted of EVs (as reported in 14 ).

| D ISCUSS I ON
The vesicles isolated from T. muris ES fit the size and shape characteristics for classification as exosomes, and previously described exosome markers (including tetraspanins, heat shock proteins and Alix) 30 were identified within these samples. Mass spectrometry analysis showed that the majority of T. muris EV proteins lack a signal peptide (68%) and that there was significant overlap between the protein content of EVs and ES (77% of EV proteins were identified in ES). This suggests that EVs may be an important mechanism by which these proteins are released into the external environment.
Similarly, Marcilla and colleagues reported significant overlap between the protein content of F. hepatica and E. caproni ES and EV samples. 25 Here, we show that vaccination with T. muris EVs can induce offer insight into why EVs make effective vaccines. 34 It has been suggested that encapsulating antigens in lipid spheres protects them from degradation and enables slow release of antigen over time. [34][35][36] In addition, Fifis and colleagues have demonstrated that 40-50 nmsized nanoparticles are preferentially taken up by DEC205 + CD40 + CD86 + murine DCs compared to larger particles of up to 2 μm in size. 37 Therefore it is reasonable to suggest that presentation of  20 We measured high levels of antiparasite IgG2a/c in all of the EV vaccinated mice ( Figure S1), and as such, found no correlation between worm burden and antiparasite IgG2a/c production. This suggests that EV vaccinated mice mount a mixed Th1/Th2 response to a low-dose T. muris infection.
The sera of EV vaccinated mice recognize a number of components that are enriched within EV samples, demonstrating that these components are antigenic. The strongest antibody response was directed towards 100, 80 and 70 kDa components. Figure 3D shows SDS-PAGE separation of the lysed EV material and mass spectrometry analysis of the protein content within these regions revealed a number of potential antigens. These include VWD and vitellogenin N and DUF1943-domain-containing protein, vacuolar protein sortingassociated protein 52, and TSP-1 domain-containing protein, which are among the most abundant EV proteins (Table S1). Eichenberger and colleagues also identified these proteins within T. muris EVs. 23 Although antibody responses may not reflect protection, the therapeutic value of related proteins has been demonstrated in other helminths, [34][35][36][37][38][39][40][41] suggesting that these proteins are major candidates for protective antigens. Future work should investigate recombinant forms of these proteins as protective antigens and explore opportunities for EVs to boost their antigenicity.

ACK N OWLED G EM ENTS
The authors would like to thank the Bio-MS core facility (University of Manchester, UK) for their assistance with mass spectrometry.