Three isolates (Hp8, BJ, and 668) of A. paragallinarum obtained from chickens in China and previously identified to be Page serovar A, B, and C, respectively (Xing et al., 1986; Chen et al., 1993; Zhang et al., 2003) were used in this study.

Chicken macrophage HD11 cell line was derived from NEWGAINBIO (HD11, CL1337, NEWGAINBIO, Wuxi, China).

Specific pathogen-free (SPF) chickens (42 days old) were obtained from the Experimental Animal House of the China Pharmaceutical University (Beijing, China). They were given free access to pelleted feed (Beijing Cooperation Medical and Pharmaceutical Company, Beijing, China). All animal work in this study met the minimum standards of animal welfare as described in the International Guiding Principles for Biomedical Research involving Animals (at https://grants.nih.gov/grants/olaw/Guiding_Principles_2012.pdf). The handling of chickens was in accordance with the Guidelines of Animal Care and Use Committee of the Institute (IACUC) under the approval of Institute of Animal Husbandry and Veterinary Medicine (Permit number: 2014-05). All efforts were made to avoid animal suffering and to improve the quality of life of the birds.

A. paragallinarum strain hp8 (Page serovar A) was grown in 10 ml of tryptic soy broth (TSB, BD, United States) supplemented with 10% chicken serum and 2.5% nicotinamide adenine dinucleotide (NAD) at 180 rpm for 15 h at 37°C. The culture was then transferred to 500 ml of supplemented TSB and incubated for 12 h. Subsequently, the whole bacterial cells were removed by centrifugation at 10,000 × g for 10 min at 4°C, and the collected supernatant was filtered using Stericup filters (pore size, 0.45 μm; Millipore Corporation, Billerica, MA, United States) to remove floating cell debris. The filtered supernatant was centrifuged by ultracentrifugation in a 45 Ti rotor (Beckman Coulter, Krefeld, Germany) at 235,000 × g for 3 h at 4°C to yield a crude OMV preparation. Then, the crude OMVs were resuspended in 2 ml HEPES buffer (50 mM HEPES, 150 mM NaCl, pH 6.8) containing 45% w/v iodixanol. Highly purified OMVs were prepared using OptiPrep™ (60% w/v iodixanol in water, Fresenius kabi Norge AS, Oslo, Norway) density gradient centrifugation. The crude OMVs preparation was placed in an Ultra Clear™, 14 ml, 14 × 95 mm tube (Beckman Coulter, Krefeld, Germany). A discontinuous iodixanol gradient was achieved by layering successive 2 ml volumes of HEPES buffer containing 40%, then 35, 30, 25, 20, and 15% w/v iodixanol. After centrifugation at 156,000 × g for 180 min at 4°C (using a SW40 Ti rotor installed in an Optima L-80XP Ultracentrifuge, Beckman-Coulter, Krefeld, Germany), the visible enriched fractions in the various layers of the density gradient were collected by pipette (in a sequence from top to bottom). Gradient fractions containing purified OMVs were identified by protein analysis using the bicinchoninic acid (BCA) protein assay kit (Thermo Scientific, United States) according to the manufacturer’s instructions. After that, the highly purified OMVs were diluted with sodium chloride-Tris-ethylenediaminetetraacetic acid (EDTA) buffer (STE; Santa Cruz, CA, United States) and centrifuged at 156,000 × g for 120 min at 4°C (using a SW40 Ti rotor, Beckman-Coulter, Krefeld, Germany). The pellet was collected and resuspended in 1 ml phosphate-buffered saline (PBS) as the final OMV preparation and stored at −20°C.

Formvar-coated copper grids were floated on 20 μl droplets of purified membrane vesicle solutions/and or purified OMVs with A. paragallinarum [a single colony, grown on tryptic soy agar (TSA, BD, United States) supplemented with 10% chicken serum and 2.5% NAD for 15 h at 37°C in candle jar] for 20 min. The grids were rinsed in water and then floated on 1% uranyl acetate for 30 s. After washing again with water, the grids were imaged on a Philips CM 100 transmission electron microscopy (TEM) at an accelerating voltage of 80 kV.

To eliminate membrane-associated nucleic acids, an issue reported by Ramon Rocha et al. (2006), in the OMV preparation, 1 U/μl of DNAse I (Thermo Fisher Scientific, CA, United States) was added to the preparation, and the mixture was incubated for 10 min at 37°C. Next, OMVs were pelleted by centrifuging at 15,000 × g for 15 min, and the supernatant was removed by micropipette aspiration. OMVs were disrupted by incubating at 70°C for 10 min with 400 μl of VL buffer [40 mM Tris-acetate, pH 7.8, 20 mM sodium acetate, 1 mM EDTA, 1% sodium dodecyl sulfate (SDS; w/v)]. The mixture was cooled to room temperature in a water bath for 3 min. To remove OMV debris from the suspension, 150 ml of 5 M sodium chloride was added, mixed by inverting the tube 30 times, and centrifuged (15,000 × g, 4°C, 10 min).

DNA was extracted from the DNase-treated OMVs by a commercial kit (Bacterial Genomic DNA Extraction Kit, Solarbio, China). A PCR that targeted the bacterial 16S ribosomal RNA (rRNA) gene was performed using the primer pair 27F (AGAGTTTGATCCTGGCTCAG) and 1492R (GGTTACCTTGTTACGACTT; Garcia-Sanchez et al., 2014). After amplification, products were purified using a commercial kit (Invitrogen PCR product purification kit, Life Technologies, United States) and were directly sequenced using an ABI PRISM BigDye Terminator V3.1 kit (Applied Biosystems, CA, United States). The sequences were analyzed and edited using Sequencing Analysis V5.2 Software (Applied Biosystems, CA, United States) and compared with already published sequences in the GenBank, NCBI, using the BLAST tool.

The purified DNA from OMVs was also sent for whole genome sequencing (Allwegene Technologies, China) on the Illumina HiSeq 4000 PE150 platform with massively parallel sequencing (MPS) Illumina Technology. Illumina PCR adapter reads and low quality reads from the paired-end and mate-pair library were filtered by a quality control step that used the provided compiling pipeline. All good-quality paired reads were assembled using the SOAP de novo software with at least 100-fold coverage¹ into a number of scaffolds. Then, the filtered reads were subjected to gap closing.

The kinetic chromogenic limulus amebocyte lysate (LAL) assay (Kinetic-QCL; Lonza) for endotoxin measurement was performed as previously described (Thorne, 2000). Briefly, 10-fold dilutions, prepared in pyrogen-free water, of the OMV preparations were made in depyrogenated borosilicate tubes. Twofold dilutions of the control standard endotoxin/LPS (Escherichia coli O55:B5, Sigma, SC, United States) were assayed to create a 12-point standard curve ranging from 0.0244 to 50.0 endotoxin units (EU)/ml. The preparations and controls were assayed in 96-well microplates, (Corning) and the absorbance change was measured at 405 nm every 30 s for 90 min using a microplate reader (Spectra Max 384 Plus, Molecular Devices, with Softmax PRO 4.0). All samples assayed by the LAL assay were diluted to yield at least four separate dose points that fell within the controlled standard endotoxin (CSE) curve. Quality assurance measures for the LAL assay were as previously described (Thorne et al., 2015).

Detection of proteins from OMVs on gel electrophoresis was carried out according to the manufacturer’s instructions. The 20-μl OMV sample was loaded on to a 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel. After electrophoresis, the gel was transferred to a staining solution containing Coomassie Blue Stain G-250. The gel was then destained with 15% (v/v) methanol and 10% (v/v) acetic acid.

The bands of interest from two hp8 OMV SDS-PAGE gel lanes (corresponding to 65 and 15 kDa) were excised, pooled, and cut into 1-ml cubes. Then, the samples were subjected to mass spectrometric analysis at the Beijing Genomics Company by using a Q Exactive mass spectrometer (Thermo Scientific, CA, United States). Only doubly and triply charged ions, with an intensity of >15 counts/s, were selected for tandem mass spectrometry (MS/MS) analysis. Data acquisition and processing were performed using MassLynx version 4.0 software (Micromass, United Kingdom). All samples were measured in triplicate.

Chicken macrophage cells (HD11) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) high glucose (Invitrogen Corporation, MA, United States) supplemented with 10% fetal serum, 100 U/ml streptomycin, and 100 mg/ml penicillin at 37°C in an atmosphere of 5% CO₂. The medium was changed three times per week, according to a standard cell culture protocol. After reaching at least 85% confluence, cells were suspended in a complete growth medium. Trypan blue (100 μl of a 0.1% solution) was added into 100 μl of the cell suspension and gently mixed. Cells were then counted with a hemocytometer, as per the manufacturer’s instructions. The cells, adjusted to 2 × 10⁶/ml, were plated in 25-cm² flasks (Corning, NY, United States) at 5 ml/flask.

HD11 cells were treated with 5 μg/ml OMVs (protein concentration) for 24 and 48 h. In addition, the control HD11 cells were treated with 1 μg/μl LPS (E. coli O55:B5, Sigma, United States) for 24 and 48 h as well. Total RNA was isolated from OMV or LPS-treated cells with TRIzol reagent (Invitrogen). RNA concentration was measured with NanoDrop One (Thermo Fisher Scientific, WI, United States) using the optical density ratio at 260 and 280 nm (OD260/OD280). Complementary DNA (cDNA) was synthesized by RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, CA, United States) according to the manufacturer’s instructions. The cDNA samples were stored at −20°C for subsequent qRT-PCR.

A quantitative real-time PCR (qRT-PCR) assay was carried out using a LightCycler® 96 SW 1.1 instrument (Roche, Germany). The reaction mixture was composed of SYBR Premix Ex Taq II (Takara, China), the relevant specific primer set (Table 1), and DNA template. β-actin was used as a housekeeping gene (De Boever et al., 2008). Quantitative PCR was performed with ABI ViiA7 Real-Time PCR System (Applied Biosystems, Carlsbad, CA, United States). The amplification program used was as follows: an initial denaturation step at 95°C for 10 min, followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 1 min and extension at 72°C for 1 min, and a final extension step at 72°C for 8 min. The results were analyzed by the CT method using the onboard data assist software (version 3.0, Applied Biosystems/Life Technologies). The data were normalized to the β-actin gene expression.

The hemagglutination (HA) activity of OMVs was determined using a standard HA microtiter assay. Briefly, twofold dilutions of OMVs in 25 μl volumes of PBS were prepared in duplicate in a U-bottomed microtiter plate. Thereafter, 25 μl of either 1% fresh or glutaraldehyde-fixed chicken red blood cells was added to each well containing OMVs plus the control wells that contained 25 μl of diluent only. The plates were incubated at room temperature for 30–60 min, and red blood cells agglutination was detected by direct visual inspection.

A mixture of OMVs and MONTANIDE™ ISA 71 VG adjuvants (SEPPIC, Shanghai, China) was emulsified at a 3:7 mass ratio (antigen:adjuvant). The vaccine, termed V-OMVs-A, contained 100 μg/ml of hp8 OMVs (protein concentration). The 42-day-old SPF chickens were randomly divided into a V-OMVs-A vaccinated group and a PBS control group (n = 30 in each group). A 0.5-ml volume of V-OMVs-A was injected into chicken by the breast muscle route in the vaccinated group. The second injection was administered at 2 weeks after the first dose. The control birds were inoculated with PBS.

Four weeks after the second immunization, all chickens from each group were challenged with one of the three serovars of A. paragallinarum (10 chickens for each serovar) – hp8 (5 × 10⁵ CFUs/0.2 ml), BJ (2 × 10⁵ CFUs/0.2 ml), and 668 (5 × 10⁵ CFUs/0.2 ml), by intranasal inoculation. Clinical signs of infectious coryza, such as mucous exudate from the nostrils, sneezing, swelling of the infraorbital sinuses, facial edema, and conjunctivitis, were recorded from the second to the eighth day postchallenge. All birds were monitored daily, and an individual score was given based on clinical signs: 0, active with no adverse clinical sign; 1, mild facial swelling and or nasal discharge; 2, moderate facial swelling and or nasal discharge, slightly weak with dropped wings weak; and 3, severe facial swelling and or nasal discharge, unresponsive with eyes closed. All chickens were then humanely euthanized, and swabs, taken from both infraorbital sinuses, were cultured on blood agar with a Staphylococcus epidermidis nurse colony (Soriano et al., 2004). The inoculated plates were incubated at 37°C with 5% CO₂. The presence of satellitic colonies, indicating the recovery of the challenge strain of A. paragallinarum, was recorded after overnight incubation. Protection was defined as the absence of clinical signs of infectious coryza and a failure to reisolate A. paragallinarum.

Four weeks after the second immunization, serum samples were collected from all chickens, and antibody titers against OMVs were tested by ELISA. A 96-well microplate (Corning) was coated with 1 μg/well of sonicated bacterial antigen prepared from hp8 and diluted with a carbonate/bicarbonate buffer (pH 9.6). The coated plates were incubated overnight at 4°C. After blocking with 5% skimmed milk for 2 h (at 37°C), a 100-μl aliquot of the serum sample (diluted 1:100 in PBS) was added into the 96-well microplate and incubated at 37°C for 1 h. Then, 100 μl of horseradish peroxidase conjugated goat antichicken IgG (IgG-HRP; Abcam, Cambridge, United States) was added as a secondary antibody, followed by incubation at 37°C for 1 h. Finally, the microplate was read at 450 nm using an EON microplate spectrophotometer (Biotek, VT, United States).

Data in the results were analyzed by the independent sample t-test and one-way ANOVA using GraphPad Prism software (GraphPad Software, Inc., San Diego, CA, United States). All results were presented as mean ± standard deviation (SD). p < 0.05 were regarded as statistically significant.

The OMVs were found to be both closely associated with bacterial cells as well as distinctly separated from the cells of A. paragallinarum. The OMVs were spherical in shape and had membranous structures with diameters varying between 20 and 300 nm (Supplementary Figure S1).

OMVs contained DNA at an estimated concentration of 124 ng/μl. The 16S rDNA PCR of the OMVs produced a product that was similar in size (1,398 bp) to the product obtained when using A. paragallinarum hp8 as template (data not shown). In addition, the 16S rRNA gene sequence of the OMV preparation (accession number MK806690) and multiple sequences of the 16S rDNA of A. paragallinarum were up to 99.86% identity with multiple sequences of A. paragallinarum 16S rDNA in GenBank.

The total DNA extracted from the OMVs was examined by agarose gel electrophoresis (data not shown). The genomic sequences of the OMVs of A. paragallinarum strain hp8 (BioSample accession: SAMN12734518) were composed of 184 contigs for 2,357,467 bp with a 40.99% G + C content. The largest contig was 98,601 bp, and the minimum contig was 263 bp. Automatic gene prediction was performed using GeneMarkS.² In total, 2,382 protein-encoding genes were predicted, with a coding percentage of 86.76%. The average gene length was 859 bp. In addition, the genes were searched against the Kyoto Encyclopedia of Genes and Genomes (KEGG; Kanehisa et al., 2004) and Clusters of Orthologous Groups (COG; Tatusov et al., 2003) databases to annotate the gene description. Based on the analysis, the functions of 1,448 genes (60.8% of the total of 2,382 genes) were predicted, and 2,016 genes (84.6%) were assigned into 23 clusters of COG functions. As well, one predicted copy of both the 16S and the 23S rRNA genes and six copies of 5S rRNA gene were detected using rRNAmmer software (Lagesen et al., 2007).

The level of endotoxin in OMVs was 211.2 EU/ml. The total concentration of OMV proteins was found to be 1.632 mg/ml. The OMV proteins appeared to consist of proteins of approximately 65 and 15 kDa by SDS-PAGE (data not shown).

Mass spectrometric analysis showed that all proteins with a confidence score of ≥21 from both the 65‐ and 15-kDa bands were associated with A. paragallinarum bacterial proteins (Tables 2 and 3). The components of the OMV proteins were mainly various functional enzymes (e.g., ATP-dependent RNA helicase), structural components (e.g., streptomycin B receptor and membrane protein), and some hypothetical proteins with unknown functions.

The 2^−ΔΔCt method was used to calculate the levels of inflammation-related genes expression. At 24 h, compared to the control group, the expression levels of IL-1β, IL-10, and inducible nitric oxide synthase (iNOS) were remarkably higher (p < 0.01) in the cells exposed to OMV and LPS, while the expression levels of IL-6 and IL-2 were considerably increased (p < 0.05) in OMV-exposed cells (Figure 1). At 48 h, all five genes were expressed at significantly higher levels (p < 0.01) in the OMV‐ or LPS-exposed groups compared to the control group. Moreover, the expression level of IL-10 and IL-lβ had significantly decreased at 48 h compared to 24 h.

No HA activity against either fresh or glutaraldehyde-fixed chicken red blood cells could be detected with the OMV sample.

The results of the vaccine trial demonstrated that all chickens in the control groups exhibited typical clinical symptoms of infectious coryza (Table 4; Figure 2). In the V-OMV-A group challenged with hp8, only three chickens showed clinical signs with the same three chickens also yielding A. paragallinarum, i.e., a protection rate of 70%. As well, the daily mean clinical signs was significantly lower for the vaccinated chickens compared with the controls (Figure 2). The protection in OMV-vaccinated birds challenged with either BJ or 668 was much lower, 10 and 30%, respectively. However, for all three serovars, the clinical signs at all daily observations were significantly lower in the vaccinates than in the controls (Figure 2). These results suggest that the immunogenicity of vesicles is providing a degree of cross-serovar protection (in terms of clinical signs) but that the ability to significantly lower colonization is serovar specific.

Serum IgG antibody were measured in OMV-vaccinated and control groups at 4 weeks after the secondary immunization (Figure 3). The IgG antibody level in the OMV-vaccinated group was significantly higher (p < 0.01) than that in control group. These results indicate that OMVs can stimulate strong humoral immunity in vaccinated chickens.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.