Evaluation of a toxoid fusion protein vaccine produced in plants to protect poultry against necrotic enteritis

Expression vector construction

A PlcC-NetB fusion protein gene was designed using codons preferred for N. benthamiana (Geyer et al., 2010). Amino acid sequences were based on GenBank accessions AAP15462.1 for PlcC and ABW71134.1 for NetB. For targeting of nascent protein to the endoplasmic reticulum (ER), the barley alpha amylase signal peptide, followed by a 6His tag, was added to the N-terminus of PlcC. A linker peptide GGSGGSGGPSGGSGG was inserted between the PlcC and NetB sequences (Fig. 1). The fusion protein nucleotide sequences are in GenBank accession numbers MH883808 and MH883809. The gene was synthesized commercially with flanking XbaI and SacI sites. The 1,402 bp XbaI-SacI fragment was ligated with geminiviral vector pBYR2eK2M (Diamos, Rosenthal & Mason, 2016) to make pBYR2eK2M-6HplcCnetB. Removal of the 417 bp BamHI fragment containing the PlcC and linker sequences yielded the ER-targeted 6His-NetB construct pBYR2eK2M-6HnetB. To make the ER-targeted 6His-PlcC vector, we PCR end-tailored the PlcC sequence to introduce a stop codon and SacI site at its 3′ end, using the primer PlcC-Sac-R (Table S1), inserting the resulting product to yield pBYR2eK2M-6HplcC. To produce the cytosol-targeted proteins, the signal peptide was removed by introducing an XbaI site and start codon 5′ of the 6His sequence by PCR with primer 6H-Xba-F and using the ER-targeting constructs above as templates. The resulting XbaI-SacI fragments were ligated with plasmid pBYR2e3K2Mc-GFP (Diamos & Mason, 2018; Huang & Mason, 2004) digested XbaI-SacI, thus yielding the vectors pBYR2e3K2Mc-c6HplcCnetB, pBYR2e3K2Mc-c6HplcC, and pBYR2e3K2Mc-c6HnetB.

Agroinfiltration of plants for protein production

Expression vectors were inserted into A. tumefaciens EHA105 via electroporation. PCR and restriction digestion of the purified plasmids were used to confirm the resulting strains. Agrobacterium strains were grown overnight at 30 °C in YENB (7.5 g Yeast Extract and eight g Nutrient Broth for one L) + 50 mg/L kanamycin and 2.5 mg/L rifampicin. Bacteria were pelleted for 10 min at 5,000 g, then resuspended in infiltration buffer (10 mM 2-(N-morpholino) ethanesulfonic acid (MES), pH 5.5 and 10 mM MgSO₄) with a final OD600 = 0.2. Suspensions were infiltrated into N. benthamiana leaves at 5- to 6-weeks of age using a syringe without needle via a small puncture (Huang & Mason, 2004). Infiltrated leaves were harvested at five days post infiltration (DPI).

Protein extraction, acid precipitation clarification, and filtration

Total protein was extracted from agroinfiltrated leaf samples using a 1:3 (w:v) ratio of extraction buffer (50 mM sodium phosphate, 200 mM sodium chloride, pH 8.1) at 4 °C. Leaf samples were homogenized for five to ten min until an even consistency was obtained using a two speed Conair™ Waring™ Laboratory Blender (Thermo Fisher Scientific, Waltham, MA, USA). Homogenized leaf tissue was stirred at high speed for 10 min at 4 °C to improve protein solubility.

Acid precipitation was performed on crude extracts after this 10-min stirring period. Note that no filtration, sedimentation, or centrifugation was performed before acid precipitation. One M phosphoric acid was added to the crude extract as it spun until a pH of 4.4–4.6 was reached, as measured by a VWR Sb70 Symphony Ph Meter (VWR, Radnor, PA, USA), Extract was stirred for 10 min at 4 °C. Two M Tris base was then added to the extract until pH reached 7.5–7.7, and the mixture was stirred for another ten min at 4 °C before being filtered through Miracloth and centrifuged at 17,000 g, 4 °C, for 30 min. The extract was then subjected to a second filtration through Miracloth resulting in a clarified extract, which was passed through a 0.45-micron filter for sterilization before metal affinity chromatography.

Three mL of Talon metal affinity resin (Clontech/Takara Bio, Mountain View, CA, USA) was spun down at 10,000 g, 4 °C, for 5 min and the supernatant removed from the bed. The resulting 1.5 mL resin bed was washed using 15 mL of extraction buffer (referred to below as wash buffer) and once again spun down at 10,000 g, 4 °C, for five min. Supernatant was discarded and the resin resuspended in 15 mL wash buffer before being added to a 20 mL column. A final wash of 20 mL was performed before the filtered extract was passed through the column. The resin was washed with 60 mL of wash buffer, and elution of bound proteins achieved with PBS + 150 mM imidazole, pH 7.4, collected in five 1 mL samples. Elution samples were dialyzed overnight in one L of sterile PBS using either a three mL or 10 mL 3,500 MWCO Slide-A-Lyzer G2 dialysis cassette (Thermo Fisher Scientific, Waltham, MA, USA), following manufacturer’s instructions. Elution concentrations were determined both before and after dialysis by an A₂₈₀ absorption using PBS and PBS + 150 mM imidazole, respectively, as blank solutions. Protein concentrations were confirmed by Coomassie staining of SDS-PAGE (below) analyzed with ImageJ (Schneider, Rasband & Eliceiri, 2012).

SDS-PAGE and Western blot

Crude extracts, clarified extracts, and eluted samples were mixed with SDS sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.02% bromophenol blue) under reducing conditions (0.5 M DTT) and samples boiled for 10 min and separated on 4–15% polyacrylamide gels (Bio-Rad, Hercules, CA, USA). Gels were stained with Coomassie stain (Bio-Rad, Hercules, CA, USA) or electroblot transferred to a PVDF membrane. PVDF membranes were blocked with 5% PBSTM (PBS with 0.05% Tween-20 and 5% nonfat dry milk) overnight at 4 °C. Membranes were then washed with PBST before incubation with a 1:10,000 dilution of rabbit anti-PlcC or rabbit anti-NetB antibodies (Kulkarni et al., 2010) in 1% PBSTM for 1 h at 37 °C. After washing a second time the membranes were incubated with a 1:10,000 dilution of goat anti-rabbit horseradish peroxidase conjugate in 1% PBSTM for 1 h at 37 °C. Bound antibody was detected with ECL reagent (Amersham Biosciences, San Francisco, CA, USA). Stripping of blots for reprobing was performed using a modified Abcam protocol. PVDF membrane was washed for 10 min in mild stripping buffer (for one L: 15 g glycine, one g SDS, 10 mL Tween20, pH 2.2) per Abcam’s recipe. PVDF membrane was then rinsed twice in 20 mL of PBS for 10 min followed by two 5-min rinse steps in 20 mL PBST. 5% PBSTM was used to block the membrane before reprobing.

Production of recombinant proteins in E. coli for immune assays

His-tagged recombinant PlcC (rPlcC) and NetB (rNetB) were cloned by PCR using primer sets Plc_pqe_f1/Plc_pqe_r2 and netB_pqe_f3/netB_pqe_r4, respectively (Table S1), using pYA5130 as the template. The resulting PCR fragments were digested with BamHI and SalI and ligated to BamHI/SalI-digested plasmid pQE30. Plasmids carrying the desired inserts were moved into E. coli strain M15(pREP4) (Qiagen, Hilden, Germany). Protein synthesis was induced by the addition of IPTG according to the manufacturer’s instructions. rPlcC and rNetB were purified by passage over a TALON affinity column (Clontech Laboratories, Inc, Mountain View, CA, USA) following the manufacturer’s recommendations. The GST-NetB protein shown in Figs. 2 and 3 was purified as previously described (Jiang et al., 2015).

Chicken experiments

All animal experiments were conducted in compliance with the Arizona State University Institutional Animal Care and Use Committee (Protocol 16-1480R) and the Animal Welfare Act, in accordance with the National Institutes of Health guide for the care and use of laboratory animals (NIH Publication No. 8023, revised 1978).

Groups of one-day-old Cornish × Rock broiler chickens were purchased from the Murray McMurray Hatchery (Webster City, IA, USA) and randomly sorted and placed in pens with pine shavings on the floor. Food and water was supplied ad libitum. For Experiments 1 and 2, groups of birds were immunized subcutaneously with 50 μg of fusion protein (Exp. 1, seven vaccinated and 11 control birds) or 100 μg of fusion protein (Exp. 2, 13 vaccinated and 10 control birds) with 50 μg of Quil-A adjuvant. Birds were given three immunizations at weekly intervals beginning at 8 days of age, with the final immunization at 22 days of age. Mock vaccinated controls were injected with adjuvant only. At 27 days of age, blood was taken for serum analysis. Three days later, when the birds were 30 days of age, birds were given an in-feed challenge with virulent C. perfringens for 5 days. The birds were scored for intestinal lesions the next day, at 35 days of age. For Experiment 3, birds were injected with 50 μg of PlcC-NetB at 7 days of age (22 vaccinated and 19 controls) and again at 12 days of age. Blood and tissues were taken from five birds/group for immunological analysis 10 days after the boost. Birds were challenged 2 weeks after the final boost, when the birds were 26 days of age.

Enzyme-linked immunosorbent assay (ELISA)

ELISAs were performed in triplicate as described Williamson & Titball (1993) to determine the IgY responses against his-tagged PlcC, his-tagged Fba, his-tagged NetB and S. Typhimurium lipopolysaccharide in chicken sera and IgA, IgY and IgM responses in intestinal washes. Biotinylated anti-chicken IgA (Alpha Diagnostic Intl. Inc., San Antonio, TX, USA), IgY (Southern Biotechnology, Birmingham, AL, USA) or IgM (Bioss, Woburn, MA, USA) antibodies diluted 1:10,000 were used to detect the various antibody isotypes.

Cellular proliferation assay

A proliferation assay was performed to evaluate cell-mediated immunity. Lymphocytes in the blood were harvested using the gentle swirl technique (Timbermont et al., 2011) and plated in quadruplicate, in a 96-well plate at 10⁵ cells/well in RPMI-1640 without phenol red. Spleens were placed through a 70 μm cell strainer to obtain single cell suspensions. Red blood cells were lysed with Red Blood Cell Lysis solution (eBioscience, San Diego, CA, USA). Splenocytes were then washed, suspended in RPMI and plated at 10⁶ cells/well. Each set of cells was incubated at 37 °C, 5% CO₂ for 72 h with or without four μg/mL of either His-Fba, His-NetB, His-PlcC, or one μg/mL PMA. Cell proliferation was measured using the Vybrant®MTT Cell Proliferation Assay Kit (Molecular Probes, Eugene, OR, USA). Mean absorbance value of antigen stimulated wells divided by mean absorbance of non-stimulated control wells was used to calculate stimulation index.

Challenge with C. perfringens

Chickens were fed an antibiotic-free starter feed containing 21% protein for 20 days, at which time the feed was switched to a high protein (28% protein), wheat-based feed containing 36% fish meal and zinc at 400 ppm (customized by Reedy Fork Farm, Elon, NC, USA) to predispose the birds to NE (Coursodon et al., 2012; Shojadoost, Vince & Prescott, 2012). Birds were challenged with virulent C. perfringens strain CP4 (Thompson et al., 2006) in feed for five days as described (Shojadoost, Vince & Prescott, 2012) with some modifications. Feed was withdrawn on the day prior to challenge for 15 h. The following day, chickens were inoculated with 1.0 mL of an overnight culture of C. perfringens CP4 grown in cooked meat medium (CMM; Difco, Detroit, MI, USA) by oral gavage. Immediately after gavage, infected feed was provided thereafter for five consecutive days. To prepare infected feed, C. perfringens was grown in CMM for 24 h at 37 °C, which then was inoculated into fluid thioglycollate medium (Difco) at a ratio of 0.3% (v/v) and incubated at 37 °C for 15 h (morning challenge) or 23 h (evening challenge). The C. perfringens culture was mixed with feed at a ratio of 1:1 (v/w). Infected feed was prepared freshly twice daily. All birds were euthanized and necropsied the day following the final challenge.

Lesion scoring

Protection against C. perfringens challenge was assessed on the basis of gross intestinal lesion scores at necropsy. On day 33, chickens were euthanized with CO₂ and their small intestines (defined here as the section between the gizzard and Meckel’s diverticulum) were examined for visible gross lesions. Intestinal lesions were scored as follows: 0 = no gross lesions; 1 = thin or friable wall or diffuse superficial but removable fibrin; 2 = focal necrosis or ulceration, or non-removable fibrin deposit, 1–5 foci; 3 = focal necrosis or ulceration, or non-removable fibrin deposit, 6–15 foci; 4 = focal necrosis or ulceration, or non-removable fibrin deposit, 16 or more foci; 5 = patches of necrosis two to three cm long; 6 = diffuse necrosis typical of field cases (Shojadoost, Vince & Prescott, 2012). Birds that died during challenge were assigned a score of 6.

Statistical analysis

All statistics were carried out using GraphPad Prism 6.0 (Graph-Pad Software, San Diego, CA, USA). Antibody titers and lesion scores were analyzed using one-way ANOVA followed by Tukey’s posttest. The values were expressed as means ± SEM, and differences were considered significant at P < 0.05.

Construction of a PlcC-NetB fusion protein

We designed a fusion protein comprising an N-terminal 6His tag, PlcC, linker GGSGGSGGPSGGSGG, and NetB (Fig. 1). The long, flexible linker was used to facilitate correct folding of the PlcC and NetB domains, and to permit unencumbered access of B-cell receptors to the important immunogenic regions. PlcC contains the C-terminal receptor-binding domain of the alpha toxin (Guillouard et al., 1997). The NetB sequence contains a mutation W262A in the rim domain (Savva et al., 2013), which substantially attentuates the toxicity of the protein in cell culture assays. We noted that PlcC and NetB contain a total of five eukaryotic Asn-linked glycosylation sites (N-X-S/T), which could result in the addition of glycans to the Asn residue when the protein is targeted to the ER in plant cells. We constructed expression vectors to target proteins either to the ER (pBYR2eK2M-6HplcCnetB, -6HplcC and -6HnetB) or to the cytosol (pBYR2e3K2Mc-c6HplcCnetB -c6HplcC, and -c6HnetB). We expected the cytosol targeted proteins to be unglycosylated (designated PlcC-netB-NG), while the ER-targeted proteins may be glycosylated at any combination of or all of the five consensus N-X-S/T sites.

Production of PlcC-NetB and PlcC-NetB-NG fusion proteins in plants

Using the methods outlined herein two different fusion proteins were produced. One, targeted to the ER, was glycosylated (PlcC-NetB). The other, targeted to the cytoplasm, was non-glycosylated (PlcC-NetB-NG). Both fusion proteins were probed with rabbit anti-PlcC and rabbit anti-NetB antibodies which yielded the same results on a western blot (Fig. 2). PlcC-NetB also has incrementally increased band sizes above the initial band which aligns with the singular PlcC-NetB-NG band at approximately 49 kDa, the expected relative size of the fusion proteins (Fig. 2). Both NetB and PlcC expressed individually were also probed and appear at the expected sizes of 34 kDa and 15 kDa, respectively. While ER and cytoplasmic targeted NetB and PlcC were produced, only representative results are shown in Fig. 2. The yield of all PlcC-NetB constructs was estimated to be ∼2 mg/g leaf mass (∼20% of total soluble protein).

The process of purification for the PlcC-NetB protein is outlined in Fig. 3. A Coomassie-stained gel (Fig. 3A) shows a wild-type, uninfiltrated control followed by crude extract which had already been acid precipitated. Note that the crude extract has highly reduced plant protein contaminants, due to the acid precipitation that caused contaminating proteins to become insoluble while the fusion proteins remain soluble. The flow-through fraction contained bands of a similar profile to the crude extract on the Coomassie, while both washes showed no detectable protein. Three elution fractions contained bands at approximately 49 kDa. Anti-NetB antibodies (Fig. 3B) reacted primarily with the standard and elution 2, while elutions 1 and 3 had visible but less prominent bands. When probed with anti-PlcC antibodies (Fig. 3C), similar bands were present in every lane except the blank and NetB standard lanes. Elution fraction bands were as strong as those seen in the crude extract lane. Flow-through and wash fractions had lighter but detectable bands when probed with anti-PlcC. Bands that reacted with anti-PlcC antibodies were also seen at approximately 49 kDa.

Serum antibody responses to fusion protein antigens

We evaluated the serum antibody responses in birds immunized with PlcC-NetB and in control birds against the glycosylated fusion protein and against non-glycosylated recombinant rPlcC and rNetB produced in bacteria. In Experiments 1 and 2, the anti-PlcC-NetB titers were significantly higher than controls (Fig. 4). The titers against rPlcC and rNetB were also significantly elevated over controls. In Experiment 3, where birds were immunized twice, only the titers against the fusion protein were significantly higher than controls, although the serum titers against rNetB were elevated.

Cellular responses against PlcC-NetB in Experiment 3

Immunization with the fusion protein led to variable cellular responses. The rPlcC-NetB fusion protein elicited weak, but significant, cellular responses against rNetB and rPlcC-NetB in splenocytes (Fig. 5). No antigen-specific stimulation was detected in lymphocytes (data not shown).

Protection against C. perfringens challenge

Protection against challenge was assessed by scoring intestinal lesions using a six-point scoring system (Shojadoost, Vince & Prescott, 2012). Immunization with the rPlcC-NetB fusion protein resulted in a significant reduction in lesion scores in all three experiments (Fig. 6). Note that in Experiment 3, birds were challenged a week earlier than in Experiments 1 and 2. The younger age of the birds may have contributed to the higher lesion scores for some birds. Two of the non-immunized birds died during challenge. This was the only experiment in which any birds died following challenge.