Application of a Novel Epitope- and Structure-Based Vaccinology-Assisted Fimbria-Toxin Multiepitope Fusion Antigen of Enterotoxigenic Escherichia coli for Development of Multivalent Vaccines against Porcine Postweaning Diarrhea

ETEC-associated postweaning diarrhea (PWD) causes significant economic losses to swine producers worldwide. Currently, there is no effective prevention against PWD. A vaccine that blocks ETEC fimbriae (K88 and F18) from attaching to host receptors and prevents enterotoxins from stimulating water hypersecretion in pig small intestinal epithelial cells can effectively protect against PWD and significantly improves pig health and well-being. The fimbria-toxin MEFA generated from this study induced neutralizing antibodies against both ETEC fimbriae and all four ETEC toxins, suggesting a great potential of this fimbria-toxin MEFA in PWD vaccine development and further supporting the general application of this novel MEFA vaccinology platform for multivalent vaccine development.

ABSTRACT

Enterotoxigenic Escherichia coli (ETEC) strains producing K88 (F4) or F18 fimbriae and enterotoxins are the predominant cause of pig postweaning diarrhea (PWD). We recently identified neutralizing epitopes of fimbriae K88 and F18, heat-labile toxin (LT), heat-stable toxins type I (STa) and type II (STb), and Shiga toxin 2e (Stx2e). In this study, we explored a novel epitope- and structure-based vaccinology platform, multiepitope fusion antigen (MEFA), for PWD vaccine development. By using an epitope substitution LT toxoid, which lacks enterotoxicity but retains immunogenicity, as the backbone to present neutralizing epitopes of two ETEC fimbriae and four toxins, we generated PWD fimbria-toxin MEFA to mimic epitope native antigenicity. We then examined MEFA protein immunogenicity and evaluated MEFA application in PWD vaccine development. Mice subcutaneously immunized with PWD MEFA protein developed strong IgG responses to K88, F18, LT, and STb and moderate responses to the toxins Stx2e and STa. Importantly, MEFA-induced antibodies inhibited adherence of K88 or F18 fimbrial bacteria to pig intestinal cells and also neutralized LT, STa, STb, and Stx2e toxicity. These results indicated that PWD fimbria-toxin MEFA induced neutralizing antibodies against an unprecedent two fimbriae and four toxins and strongly suggested a potential application of this MEFA protein in developing a broadly protective PWD vaccine.

IMPORTANCE ETEC-associated postweaning diarrhea (PWD) causes significant economic losses to swine producers worldwide. Currently, there is no effective prevention against PWD. A vaccine that blocks ETEC fimbriae (K88 and F18) from attaching to host receptors and prevents enterotoxins from stimulating water hypersecretion in pig small intestinal epithelial cells can effectively protect against PWD and significantly improves pig health and well-being. The fimbria-toxin MEFA generated from this study induced neutralizing antibodies against both ETEC fimbriae and all four ETEC toxins, suggesting a great potential of this fimbria-toxin MEFA in PWD vaccine development and further supporting the general application of this novel MEFA vaccinology platform for multivalent vaccine development.

INTRODUCTION

The predominant cause of postweaning diarrhea (PWD) in pigs is enterotoxigenic Escherichia coli (ETEC) strains that produce K88 (F4) or F18 fimbriae and enterotoxins, including heat-labile toxin (LT), heat-stable toxin type I (STa) and heat-stable toxin type II (STb) (1–4). Shiga toxin type 2e (Stx2e) is also prevalent in ETEC strains isolated from pigs with PWD (5). PWD may result in weight loss, slow weight gain, and acute death and causes significant economic losses to swine producers worldwide (4, 6, 7). Vaccination has been considered an effective prevention approach against PWD (2, 7). Indeed, vaccination of pregnant sows with extracted ETEC fimbriae and LT B subunit (LT_B) proteins has largely prevented diarrhea in neonatal piglets through the provision of passive immunity in the colostrum and milk (8). However, the development of broadly effective vaccines to protect recently weaned pigs from PWD has not yet been achieved.

The approach that uses an avirulent strain to express K88 fimbria and/or F18 fimbriae to develop PWD vaccines is unlikely to be broadly protective. First, products developed from this method do not carry antigens that induce antitoxin immunity and thus cannot protect against ETEC toxins, which are the virulence determinants of PWD. Indeed, since it is the toxins produced by ETEC strains that stimulate fluid and water hypersecretion in pig small intestinal epithelial cells to cause PWD, effective PWD vaccines need to target key toxins as well (1, 2, 7). Second, vaccine products carrying F18 fimbriae are not expected to induce protective anti-F18 antibodies and thus are not protective against F18 fimbrial ETEC strains, which are associated with over one-third of PWD cases (5). F18 fimbriae, either physically extracted from ETEC strains or expressed on the surface of an avirulent E. coli strain, were not effective in inducing protective antibodies against F18 ETEC infection (9, 10). It also has been suggested that a vaccine that induces protective antiadhesin and antitoxin immunity would be more effective in preventing ETEC diarrhea (11). Unfortunately, none of the PWD vaccine products licensed regionally or vaccine candidates currently under development induce protective immunity against F18 fimbriae, nor LT and ST enterotoxins (2, 7).

Progress has been made in recent years to overcome challenges in PWD vaccine development. We demonstrated that toxoid or toxin epitope fusion proteins were safe, immunogenic, and protective against ETEC enterotoxicity and cytotoxicity (12–17). Additionally, by applying multiepitope fusion antigen (MEFA), a novel epitope- and structure-based vaccinology platform (18, 19), we were able to use LT toxoid as a backbone immunogen to present STa, STb, and Stx2e toxoids or peptides, constructed a toxin MEFA to mimic toxoid and epitope native antigenicity, and further demonstrated that this toxin MEFA induced antibodies against four ETEC toxins (20). We also showed that the antigenic domains of K88 major subunit FaeG and F18 adhesin subunit FedF carried by an LT toxoid induced antibodies that inhibit adherence of K88 or F18 fimbrial ETEC and protect pigs against K88-fimbrial ETEC diarrhea (21, 22). More recently, we mapped and identified the neutralizing epitopes of K88 fimbrial FaeG subunit (23), the F18 fimbrial FedF subunit (24), and the LT_A subunit (17).

In this study, we applied the novel MEFA vaccinology platform to construct a PWD fimbria-toxin MEFA (PWD MEFA) by using the LT_A1 domain as the backbone to host an STa toxoid, an STb peptide, and neutralizing epitopes of Stx2e and K88 and F18 fimbrial adhesive subunits. Mice were immunized with this PWD MEFA protein, and toxin- and fimbria-specific immunogenicity was determined. Furthermore, we measured MEFA-induced antibodies for neutralization activities against toxin enterotoxicity or cytotoxicity and adherence by K88⁺ and F18⁺ E. coli strains and assessed the potential application of this MEFA in PWD vaccine development.

RESULTS

The PWD MEFA A1 domain carried epitopes of K88 and F18 fimbriae and four ETEC toxins and expressed recombinant MEFA proteins.

Four neutralizing fimbrial epitopes, two from F18 minor subunit adhesin FedF (QPDATGSWYD and IPSSSGTLTCQAGT) (24) and two from K88 major subunit adhesin FaeG (GRTKEAFATP and PMKNAGGTKVGSVKVN) (23) representing K88 and F18 antigens, and four toxin epitopes, KKDLCEHY of STb toxin, QSYVSSLN of the Stx2e A subunit, and two copies of the STa toxoid (STa_N11S) epitope domain CCELCCSPACAGCY (the underlined amino acid is mutated) were used to replace eight surface-exposed but less antigenic epitopes at the A1 domain of the LT_A subunit to construct the PWD fimbria-toxin MEFA (PWD MEFA) gene. Two neutralizing LT epitopes, DSRPPDEIKRSGG and SPHPYEQEVSA (17), were retained at the LT_A1 backbone (Fig. 1A). Protein three-dimensional (3D) modeling (PyMOL) predicted that all fimbria and toxin epitopes on the A1 domain protein were exposed at the protein surface (Fig. 1B and C). The A1 fragment of PWD MEFA gene was synthesized, cloned into vector pET28α, and expressed as PWD MEFA A1 domain protein. After fusing the PWD MEFA A1 gene to the LT_A2 domain and LT_B subunit nucleotides in a single open reading frame, the PWD MEFA gene was constructed, and the gene was verified by DNA sequencing. After the PWD MEFA gene was cloned into pET28α and expressed by the E. coli BL21(DE3) strain, the protein was extracted. Fimbria and toxin epitopes were confirmed to be surface exposed (Fig. 1F and G) and antigenic in silico.

FIG 1.

Illustration of PWD MEFA A1 domain gene construction and PyMOL protein modeling of the A1 domain of PWD MEFA and PWD MEFA proteins. (A) Schematic illustration of the A1 domain of the PWD fimbria-toxoid MEFA gene. Nucleotides encoding two K88 epitopes (GRTKEAFATP and PMKNAGGTKVGSVKVN), two F18 epitopes (IPSSSGTLTCQAGT and QPDATGSWYD), the Stx2e A subunit epitope (QSYVSSLN), and two copies of the STa toxoid STa_N12S domain (CCELCCSPACAGCY) and STb epitope (KKDLCEHY) were embedded in the A1 segment of LT toxoid LT_R192G by replacing LT_A1 epitope nucleotides. (B, C) 3D modeling of the A1 domain of PWD MEFA protein (front and back views). (D, E) Secondary structure of the A1 domain of PWD MEFA protein (front and back views). (F, G) 3D modeling of the protein structure of PWD MEFA (A1 domain with epitope substitution, A2 domain, and LT_B peptide in a single peptide [front and back views]). (H, I) Secondary structure of the PWD MEFA protein (A1 domain, A2 domain, and LT_B peptide in a single peptide [front and back views]). Epitopes, as well as the A2 domain and LT_B peptide, are in different colors.

The recombinant PWD MEFA A1 domain protein and the PWD MEFA protein were 28.9 and 43.3 kDa, respectively (Fig. 2A). Both proteins were recognized by anti-K88, anti-F18, anti-CT, anti-Sta, and anti-Stx2e antisera (Fig. 2B to F).

FIG 2.

Extraction and detection of the PWD MEFA A1 domain and PWD MEFA proteins. (A) SDS-PAGE Coomassie blue staining of the PWD MEFA A1 domain and PWD MEFA. (B) Western blot with anti-CT rabbit serum (1:3,000 [Sigma]; homologue to anti-LT antiserum). (C) Western blot with anti-Stx2e anti-mouse serum (1:1,500). (D) Western blot with anti-STa rabbit serum (1:2,000). (E) Western blot with anti-FedF (F18) mouse serum (1:1,500). (F) Western blot with anti-FaeG (K88) mouse serum (1:1,500). IRDye-labeled goat anti-mouse or anti-rabbit IgG (1:5,000 [LI-COR]) was used as the secondary antibody. M, protein marker (in kilodaltons [Precision Plus Protein prestained standards; Bio-Rad]). Arrows indicate the A1 domain (lane 1) and entire PWD MEFA (lane 2) proteins; total inclusion body proteins from the E. coli BL21(DE3) host strain were used as the control, indicated by “(-).”

Mice subcutaneously immunized with PWD MEFA, with or without dmLT adjuvant, developed IgG antibodies to each fimbria and toxin.

Mice subcutaneously (s.c.) immunized with PWD MEFA protein alone or PWD MEFA protein with adjuvant double mutant LT (dmLT [LT_{R192G L211A} of ETEC]) developed IgG antibodies to all target ETEC fimbriae and toxins (Fig. 3). Serum IgG titers (log₁₀) to K88, F18, cholera toxin (CT [an LT homologue]), STb, Stx2e, and STa in the mice immunized with PWD MEFA alone were 2.9 ± 0.28, 2.5 ± 0.30, 3.1 ± 0.62, 3.0 ± 0.34, 2.0 ± 0.17, and 1.4 ± 0.14, respectively. Mice immunized with PWD MEFA adjuvanted with dmLT (PWD MEFA+dmLT) developed comparable IgG titers to STb (3.3 ± 0.32) and STa (1.5 ± 0.26), but showed significantly greater serum IgG titers to K88 (3.5 ± 0.37), F18 (2.9 ± 0.38), CT (3.6 ± 0.37), and Stx2e (2.4 ± 0.13), compared to the mice immunized with PWD MEFA only (P < 0.01). No antifimbria or antitoxin IgG antibody responses were detected in the serum samples from the control mice or mice prior to immunization.

FIG 3.

Mouse IgG antibody titers (log₁₀) specific to ETEC fimbriae and toxins associated with PWD. Serum samples from each mouse in the group immunized with PWD MEFA protein with dmLT adjuvant (○), with PWD MWFA protein without dmLT (■), or PBS as the control (▲) were 2-fold serially diluted and incubated in ELISA plate wells coated with CT (Sigma [100 ng per well]), F18 fimbriae (100 ng per well), K88 fimbriae (100 ng per well), MBP-STb recombinant protein (100 ng per well), MPB-Stx2e recombinant protein (100 ng per well), or STa-ovalbumin conjugate (10 ng per well). The mean titer and standard deviation in each group are indicated by bars.

PWD MEFA-induced antibodies significantly inhibited adherence of F18⁺ and K88⁺ E. coli to porcine intestinal cell line IPEC-J2.

The serum samples from the mice immunized with PWD MEFA showed significant inhibition activities against adherence of F18⁺ E. coli and K88⁺ ETEC cells to pig small intestine cells (Fig. 4A and B). F18⁺ E. coli strain 8516, after incubation with the serum samples from the mice immunized with PWD MEFA, PWD MEFA with dmLT adjuvant, or F18 FedF adhesin subunit (as a positive control [from a separate study]) showed only 43.4 ± 4.7, 46.8 ± 9.8, and 42.6% ± 4.3% attachment to porcine intestinal cell line IPEC-J2, compared to attachment by the same bacteria incubated with control mouse serum (100%; P < 0.001). Mouse serum samples from three immunization groups (PWD MEFA, PWD MEFA+dmLT, and F18 FedF), however, showed no significant differences in inhibition against adherence to IPEC-J2 cells from strain 8516.

FIG 4.

Mouse serum antibody in vitro neutralization activities against adherence of K88 fimbrial or F18 fimbrial bacteria and enterotoxicity of STa and CT toxins. (A) Serum samples from mice immunized with PWD MEFA, PWD MEFA adjuvanted with dmLT, F18 FedF recombinant protein (as the positive control), or PBS as the control in adherence inhibition against F18 fimbrial bacteria 8516 to IPEC-J2 pig intestinal cells. (B) Serum samples from mice immunized with PWD MEFA, PWD MEFA adjuvanted with dmLT, K88 FaeG recombinant protein (as the positive control), or PBS as the control in adherence inhibition against K88⁺ ETEC strain 3030-2 to IPEC-J2 cells. The numbers of adherent bacteria (CFU) were converted to percentages, with CFU from cells treated with the control serum at 100%. (C) Mouse serum samples from two immunized groups (PWD MEFA and PWD MEFA with dmLT adjuvant) or the control in prevention of STa toxin from stimulation of cGMP in T-84 cells to show antibody neutralization activity against STa enterotoxicity. T-84 cells incubated with 2 ng STa toxin premixed with PBS, control mouse serum, or serum from each immunized group were measured for intracellular cGMP levels (pmol/ml) by using a cGMP EIA kit (Enzo Life). (D) Mouse serum samples from the two immunized groups (PWD MEFA and PWD MEFA with dmLT adjuvant) or the control in prevention of CT (LT homologue) from stimulation of cAMP in T-84 cells to show antibody neutralization activity against CT enterotoxicity. T-84 cells incubated with 10 ng CT toxin premixed with PBS, control mouse serum, or each immunized mouse serum were measured for intracellular cAMP levels (pmol/ml) by using a cAMP EIA kit (Enzo Life). ***, P < 0.001; **, P < 0.01.

K88⁺ ETEC wild-type strain 3030-2, after incubation with the serum samples from the mice immunized with PWD MEFA (41.9% ± 12.3%) or PWD MEFA with dmLT adjuvant (42.8% ± 11.8%), showed a significant reduction (P < 0.001) in adherence to IPEC-J2 cells compared to the adherence from incubation with the control serum samples (100%). When incubated with the serum from mice immunized with K88 FaeG subunit protein (as a positive control [from a separate study]), nearly 80% reduction in adherence to IPEC-J2 by 3030-2 was observed (20.6% ± 6.1%). This level of reduction was significantly greater than that afforded by serum samples from the two immunization groups (P < 0.01).

PWD MEFA-induced antibodies neutralized CT and STa enterotoxicity.

Cholera toxin (CT), a homologue of LT, and STa toxin had significantly reduced stimulation of intracellular cyclic AMP (cAMP) or cyclic GMP (cGMP), after incubation with serum samples from mice immunized with PWD MEFA (Fig. 4C and D). STa toxin (2 ng) incubated with phosphate-buffered saline (PBS), control mouse sera, sera from mice immunized with PWD MEFA or PWD MEFA plus dmLT adjuvant stimulated intracellular cGMP levels to 9.1 ± 0.40, 9.1 ± 0.40, 6.5 ± 0.77, and 5.6 ± 0.10 pmol/ml, respectively, in T-84 cells. The STa-stimulated cGMP levels in cells incubated with the serum samples from two immunized mouse groups were not significantly different (P > 0.05), but were significantly lower than that of cells incubated with control mouse sera (P < 0.01).

Ten nanograms of CT, after incubation with PBS, control mouse serum samples, the serum of the group immunized with PWD MEFA, or the serum from mice immunized with PWD MEFA adjuvanted with dmLT, stimulated cAMP in T-84 cells to 8.0 ± 1.22, 4.3 ± 0.32, 0.83 ± 0.26, and 1.1 ± 0.40 pmol/ml, respectively. The CT-stimulated cAMP levels in cells incubated with mouse sera from the two immunized groups were significantly lower (P < 0.01) than that in cells incubated with the control mouse sera.

PWD MEFA-induced antibodies neutralized STb and Stx2e cytotoxicity.

STb and Stx2e cytotoxicity in Vero cells was reduced after treatment with sera from mice immunized with PWD MEFA with or without dmLT adjuvant (Fig. 5). Filtrates of STb⁺ E. coli strain 8020 (300 μl in a final volume of 700 μl) or Stx2e⁺ E. coli strain 9168 (100 μl in a final volume of 1,000 μl) added to Vero cells resulted in characteristic cytotoxic activity that was measured as units of 50% cell death or detachment (CD₅₀). Addition of 150 μl control mouse serum (at a titer of 1:4.7 [700/150 = 4.7]) to the STb⁺ bacterial filtrates or 18.8 μl control mouse serum (at a titer of 1:53 [1,000/18.8 = 53]) to the Stx2e⁺ filtrates had no measurable effect on the respective CD₅₀ in Vero cells. In contrast, the addition of 150 μl (1:4.7 [700/150]), 25 μl (1:28 [700/25]), or even 10 μl (1:70 [700/10]) of sera from the mice immunized with PWD MEFA with or without dmLT adjuvant resulted in neutralization of STb⁺ filtrates’ (300 μl) CD₅₀ activity. Treatment with 18.8 μl (1:53 [1,000/18.8]), 6.3 μl (1:159 [1,000/6.3]), or 3.0 μl (1:333 [1,000/3]) of sera from the immunized groups neutralized all measurable cytotoxic activity by 100 μl Stx2e⁺ bacterial filtrates (Fig. 5).

FIG 5.

Mouse serum antibody in vitro neutralization activity against STb and Stx2e toxin cytotoxicity from a Vero cell assay. Vero cells incubated with 300 μl filtrates of STb⁺ strain 8020 or 100 μl filtrates of Stx2e⁺ strain 9168 showed cytotoxicity. (Top row) Normal Vero cells, cells treated with 300 μl 8020 filtrates, 300 μl 8020 filtrates premixed with 150 μl control mouse serum (1:4.7; in a total volume of 700 μl), 300 μl 8020 filtrates premixed with 150 μl (1:4.7), 25 μl (1:28), or 10 μl (1:70) mouse serum of the group immunized with PWD MEFA with or without dmLT adjuvant. (Bottom row) Normal Vero cells, cells treated with 100 μl 9168 filtrates, 100 μl 9168 filtrates premixed with 18.8 μl control mouse serum (1:53 in a total volume of 1,000 μl), 100 μl 9168 filtrates premixed with 18.8 μl (1:53), 6.3 μl (1:159), or 3 μl (1:333) mouse serum of the group immunized with PWD MEFA with or without dmLT adjuvant.

DISCUSSION

This study demonstrates that the neutralizing epitopes from ETEC virulence determinants in pig PWD can be integrated into a single immunogen using the epitope- and structure-based vaccine technology and that the resultant PWD fimbria-toxin MEFA protein induced the formation of antibodies that protected against adherence of both K88 and F18 fimbriae and enterotoxicity or cytotoxicity of all four ETEC toxins, namely, LT, STa, STb, and Stx2e. To be the best of our knowledge, this is the first report of an antigen or vaccine candidate inducing protective antibodies against all ETEC virulence factors associated with pig PWD. Since K88/LT/STb and F18/STa/STb/Stx2e are the predominant ETEC pathotypes causing PWD in weaned pigs (5), a vaccine inducing protective immunity against K88 and F18 fimbriae and toxins LT, STa, STb, and Stx2e would be broadly effective against pig PWD. The current study mainly focused on the creation of PWD fimbria-toxin MEFA protein, characterization of PWD MEFA immunogenicity, measurement of MEFA-induced antibody in vitro protection against ETEC fimbrial adherence and toxin intracellular toxic activities, and assessment of the potential application of this MEFA antigen for PWD vaccine development. Future pig immunization and challenge studies are needed to verify the efficacy of MEFA-induced immunity for protection against PWD. We suggest that a host strain or a vector system that can effectively express and secrete PWD MEFA protein onto the outer membrane and effect induction of small-intestinal mucosal immunity is needed for optimizing the vaccine format, considering parental vaccines are not desirable for young livestock animals due to concerns about cost effectiveness and the need for adjuvants and booster administration. This study nevertheless provides important data on the characterization of an immunogen toward the development of an effective PWD vaccine.

While the fimbria-toxin MEFA-induced IgG responses to the two fimbriae and LT and STb toxins were strong, only moderate responses to STa and Stx2e were observed. A lower antibody response to STa was somewhat expected since the 18-amino-acid porcine-type ETEC STa or STa toxoids are poorly immunogenic in nature (12). An early study showed that a toxin MEFA carrying STa toxoid STa_P12F (NTFYCCELCCNFACAGCY) induced low to mild anti-STa IgG titers of 0.45 ± 0.21 log₁₀ from a toxin MEFA that carried one STa_P12F and 0.87 ± 0.66 log₁₀ when three copies of STa_P12F were carried (20). Since human-type ETEC STa toxoid hSTa_N12S (equivalent to pig-type pSTa_N11S) was identified as the best at inducing neutralizing anti-STa antibodies after genetic fusion to carrier protein LT monomeric peptide LT_{R192G L211A} (16), the pig-type STa toxoid STa_N11S domain CCELCCSPACAGCY (which was thought to maintain full-length STa_N11S biological and perhaps antigenic activities) was selected as the STa antigen for PWD MEFA construction in this study. The current PWD MEFA carried two copies of the STa_N11S domain peptide and induced anti-STa IgG titers of 1.5 ± 0.26, which were greater than those of the toxin MEFA from the previous study (20). Neutralizing activity against STa enterotoxicity from PWD MEFA-derived antibodies, however, was not improved based on the cGMP EIA data (Fig. 4C). The lower neutralizing activity against STa enterotoxicity from PWD MEFA-induced antibodies may be caused by the shortened STa toxoid peptide in the PWD MEFA immunogen. Only the 14-amino-acid domain, and not the full-length molecule pSTa_N11S (18 amino acids), was used for PWD MEFA construction. The 14-amino-acid peptide is recognized as the STa toxicity domain, which maintains STa biological activities. However, the shorter length of the STa peptide could affect its proper folding and compromises antigenic integrity, thus resulting in anti-STa antibodies with poor neutralizing activity. Future studies using a full-length STa_N11S toxoid and possibly screening porcine-type STa toxoids to identify the most neutralizing toxoid for PWD MEFA construction or further optimizing STa toxoid position may help to improve MEFA-induced antibody neutralization activity against STa enterotoxicity. On the other hand, although PWD MEFA induced moderate IgG responses to anti-Stx2e, these antibodies showed great neutralizing activity against Stx2e cytotoxicity.

Results from this study, perhaps more importantly, add more supportive evidence for the epitope- and structure-based MEFA vaccinology platform and reinforce the application of MEFA technology in construction of multivalent immunogens for development of vaccines broadly protective against heterogeneous pathogenic strains. Immunologic heterogeneity among pathogenic strains or virulence determinants is a key obstacle in vaccine development. A few strategies have been attempted to overcome the heterogeneity challenge and to develop broadly protective vaccines against heterogenous strains. Those include a cocktail vaccine approach that combines a few killed or live attenuated strains to broaden vaccine coverage, application of conservative antigens that are shared among pathogenic strains but not commensal strains as vaccine immunogens, and a genetic fusion strategy or further epitope- and structure-based multiepitope fusion antigen (MEFA) vaccinology to genetically fuse multiple antigens together. MEFA integrates neutralizing epitopes from heterogeneous strains or virulence determinants into a single immunogen and mimics epitope native antigenicity for a multivalent antigen to induce broadly protective immunity. Assisted by protein modeling and molecular dynamics simulation, MEFA identifies a backbone immunogen—ideally a nontoxic and strongly immunogenic (preferably also adjuvantic) virulence determinant that possesses multiple well-separated continuous epitopes and stable secondary structure—and replaces backbone epitopes with neutralizing epitopes from heterogeneous virulence factors for broad immunogenicity (19).

By applying this MEFA technology, we generated human ETEC adhesin MEFAs, including colonization factor antigen CFA/I/II/IV MEFA and CFA adhesin tip MEFA, and ETEC adhesin-toxoid MEFA and demonstrated that these MEFAs induced broadly protective immunity. CFA/I/II/IV MEFA used strongly immunogenic CFA/I major subunit CfaB as the backbone to present neutralizing epitopes of CFA/II (CS1 to CS3) and CFA/IV (CS4 to CS6) and induced antibodies that inhibited adherence of E. coli or ETEC bacteria expressing any of the seven most important ETEC adhesins associated with diarrhea in humans (18). CFA adhesin tip MEFA with CFA/I tip adhesin subunit CfaE as the backbone to present neutralizing epitopes of nine ETEC adhesins (CFA/I, CS1 to CS6, CS21, and EtpA) induced protective antibodies against adherence of all nine human ETEC adhesins (25). CFA-toxin MEFA, which used LT monomer (one LT_A subunit and one LT_B subunit as a single peptide) as the backbone and had epitopes and a peptide on the A1 domain replaced by STa toxoid STa_N12S and CFA/I/II/IV MEFA, induced antibodies that inhibited adherence of all seven ETEC adhesins and neutralized enterotoxicity of both ETEC toxins, but more importantly protected against ETEC diarrhea in a pig challenge model (26). More recently, we constructed MEFAs that combined epitopes from the virulence factors of Vibrio cholerae and Shigella and found Vibrio and Shigella MEFAs induced antibodies to broadly inhibit adherence of different Vibrio subtypes or adherence and invasion of Shigella spp. and serotypes (data unpublished). These MEFA proteins, alone or combined with ETEC MEFAs, are under investigation for potential development of broadly protective vaccines against human diarrhea by ETEC, Vibrio cholerae, and/or Shigella spp.

Unlike a cocktail whole-cell vaccine consisting of a mixture of strains and thereby carrying excessive somatic antigens, including harmful lipopolysaccharides (LPS), an MEFA vaccine is composed of a single or a few recombinant MEFA proteins without somatic proteins or LPS. Excessive somatic antigens potentially cause host immune responses to deviate from targeting to virulence factors and cause host immune frustration, leading to low specific immune responses and subsequently poor efficacy against infections. Since an MEFA-based vaccine does not carry somatic antigens, it is less likely to cause associated side effects in particularly very young vaccinees. It instead induces strong host immune responses specific to virulence determinants, thereby providing better protection. Future efficacy studies on a vaccine product based on this PWD MEFA against pig PWD will allow us to better assess the application of the MEFA vaccinology platform in multivalent vaccine development.

MATERIALS AND METHODS

Bacteria and plasmids.

Table 1 lists the bacterial strains and plasmids used in this study. Recombinant E. coli strain 8752 expressing monomeric toxoid fusion LT-STa (15) was used as the template for PWD MEFA A1 domain or PWD MEFA gene construction. Wild-type porcine E. coli strains 3030-2 (K88ac/LT/STb) (27) and 8516 (F18ac) (5) were used as the K88⁺ and F18⁺ bacteria in antibody adherence inhibition assays. STb recombinant E. coli strain 8020 (27) and wild-type Stx2e⁺ strain 9168 (F18/Stx2e) (5) were used in preparation of bacterial filtrates for Vero cell cytotoxicity assays. Recombinant E. coli strains 9301 and 9302 (20) expressing maltose binding protein (MBP)-STb and MBP-Stx2e proteins were used to produce enzyme-linked immunosorbent assay (ELISA) coating antigens to titrate anti-STb and anti-Stx2e antibody responses. Vector pET28α (Novagen, Madison, WI) and E. coli strain BL21 codonPlus (DE3) (Agilent Technologies, Santa Clara, CA) were used for gene cloning and recombinant protein expression.

TABLE 1.

E. coli bacterial strains and plasmids used in the study

Strain or plasmid	Relevant properties	Source or reference
Strains
BL21(DE3)	huA2 Δ(argF-lacZ)U169 phoA glnV44 ϕ80Δ(lacZ)M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17	GE Healthcare
9703	A1 domain of PWD MEFA synthesized in pUC57/DH5α	This study
9715	A1 domain of PWD MEFA in pET28α/DH5α	This study
9716	PWD MEFA in pET28α/DH5α	This study
9718	A1 domain of PWD MEFA in pET28α/BL21	This study
9719	PWD MEFA in pET28α/BL21	This study
8752	STaP13F-LTR192G fusion in pET28α/BL21	15
3030-2	K88 fimbrial pig ETEC wild-type strain, K88ac/LT/STb	13
8516	F18 fimbrial pig E. coli field isolate, F18ac	5
8020	pRAS1/1826-2, K88/STb	27
9168	Pig E. coli field isolate 04-13812, F18/Stx2e	5
9301	MBP-STb fusion in pMAL-p5X/DH5α	20
9302	MBP-Stx2eA fusion in pMAL-p5X/DH5α	20
Plasmids
pET28α		Novagen
pRAS1	STb gene in pBR322	27

Construction of the PWD MEFA gene.

The A1 domain of the LT_A subunit gene (eltA) was used as the backbone for embedding nucleotide segments coding the neutralizing epitopes of K88 FaeG, F18 FedF, STa, STb, and Stx2e identified from previous studies (20, 23, 24). With the two best neutralizing epitopes of LT_A1 kept (17), the positions of eight other epitopes in the A1 domain were subjected to substitution with the above-mentioned fimbrial and toxin epitopes to construct the PWD MEFA A1 domain gene. One copy of the STb or Stx2e epitope was included based on the results of a previous study in which a toxoid MEFA antigen that carried one copy of STb and one copy of Stx2e epitope induced protective antibodies against STb and Stx2e cytotoxicity as well as diarrhea in pigs challenged with an ETEC strain producing LT and STb toxins (20). A three-dimensional protein modeling program (28–31) and the PyMOL molecular graphics system (Schrödinger LLC, New York City, NY) were used to optimize epitope substitution and protein structure and in silico antigenicity. The optimized chimeric PWD MEFA A1 DNA nucleotide fragment (with substitutions of epitope-coded nucleotides) was synthesized by GenScript (Piscataway, NJ) and initially cloned in vector pUC57 and then pET28α. This chimeric A1 fragment was subsequently fused to the LT_A2 domain and LT_B subunit gene (eltB) to form a single open reading frame for the PWD MEFA gene by using splicing overlap extension (SOE) PCR with specifically designed primers (Table 2). A PWD MEFA gene encoding a single protein (PWD MEFA) was cloned into pET28α.

TABLE 2.

PCR primers used in the study to generate the A1 domain of PWD MEFA and PWD MEFA genes

Primer	Sequence (5′→3′)	Amplified region
PWD-MEFA-F	CGGGCTAGCATGAAAAATATAACTTTC	Upstream of A1 domain of PWD MEFA, with NheI site
PWD-MEFA-R	TTACGGCCGCTAGTTTTCCATACTGAT	Downstream of LTB gene, with EagI site
PWD-MEFA-A1-L	TCATTACAAGTATCACCTGTAATTGTTCTTGAATAATTTTCACAC	Overlapping LTA1 domain with A2 domain
PWD-MEFA-A2-R	AAATTATTCAAGAACAATTACAGGTGATACTTGTAATGAGGAGAC	Overlapping LTA1 domain with A2 domain
PWD-MEFA-A1-F	CGGGCTAGCCCGATGAAAAACATCACCTTTATC	Upstream of A1 domain of PWD MEFA, with NheI site
PWD-MEFA-A-R	TTACGGCCGGAAGATGGTACGGCTGTAGTTCTC	Downstream of A1 domain of PWD MEFA, with EagI site

Expression and extraction of PWD MEFA proteins.

E. coli BL21(DE3) transformed with plasmids carrying the A1 domain or the entire PWD MEFA gene was used for protein expression and extraction as described previously (17, 20, 32). Following overnight culture, a single colony from a Luria-Bertani (LB) agar plate was picked and subcultured overnight in 5 ml LB broth supplemented with kanamycin (30 μg/ml) at 37°C. From this subculture, a 3-ml aliquot was transferred into 300 ml fresh LB broth and cultured at 37°C. Bacteria were induced with 30 μM isopropyl-β-d-thiogalactopyranoside (IPTG; Sigma, St. Louis, MO) after the optical density at 600 nm (OD₆₀₀) reached 0.6 and grown for an additional 4 h. Bacteria harvested by centrifugation (5,000 × g for 10 min) were used for inclusion body protein extraction with bacterial protein extraction reagent (B-PER; Thermo Fisher Scientific, Rochester, NY) at 4 ml per g bacterial pellet. Extracted inclusion body proteins were washed with PBS, solubilized, and refolded with solubilization buffer {50 mM CAPS [3-(cyclohexylamino)propanesulfonic acid] [pH 11.0]} supplemented with 0.3% N-lauroylsarcosine and 1 mM dithiothreitol (DTT), and dialyzed at 4°C for 24 h (with 3× buffer exchanges). Refolded proteins were examined by 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with Coomassie blue staining and Western blotting with anti-CT rabbit serum (1:3,000), anti-STa rabbit serum (1:2,000), anti-Stx2e mouse serum (1:1,500), and anti-K88 and anti-F18 mouse antisera (1:1,500), with total inclusion body proteins of BL21(DE3) used as the control. No anti-STb Western blotting was carried out due to the lack of anti-STb antiserum.

Mouse subcutaneous immunization with PWD MEFA protein and mouse antifimbria and antitoxin antibody titration.

Two groups of 8-week-old BALB/c female mice (10 mice per group) were injected subcutaneously (s.c.) with 40 μg PWD MEFA protein (in 40 μl), with or without 1 μg adjuvant double mutant LT (dmLT; LT_{R192G L211A}). A third group of 10 mice s.c. injected with 40 μl PBS was used as the negative control. Mice received 2 boosters with the same dose of the primer in an interval of 2 weeks. All mice were euthanized 2 weeks after the second booster. The mouse immunization study protocol (IACUC no. 4056) was approved by the Kansas State University Institutional Animal Care and Use Committee.

Serum samples collected from each mouse before the immunization and 2 weeks after the final booster were examined for antibodies specific to K88 and F18 fimbriae and toxins (LT, STa, STb, and Stx2e) in ELISAs as described previously (20, 23, 24). Briefly, each well of 96-well Immulon 2HB plates (Thermo Fisher Scientific) was coated with 100 ng K88 or F18 fimbria, CT (Sigma), recombinant protein MBP-STx2e or MBP-STb, or 10 ng STa-ovalbumin conjugate at 37°C for 1 h, followed by overnight incubation at 4°C. After treatment with 10% nonfat milk (in PBS) to block noncoated spots and 3 washes with PBST (PBS with 0.05% Tween 20), the plates were incubated with 2-fold dilutions (1:400 to 1:256,000) of mouse serum. Plates were washed with PBST and incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:5,000; Sigma). OD₆₅₀ values were read after exposure to TMB (3,3′,5,5′-tetramethylbenzidine) Microwell peroxidase substrate (KPL, Gaithersburg, MD) and converted into antibody titers (in log₁₀) as previously described (20, 23, 24).

Mouse serum antibody neutralization assays against K88 and F18 fimbrial adherence to porcine intestinal cell line.

Pig intestinal cell line IPEC-J2, K88⁺ strain 3030-2, and F18⁺ strain 8516 were used to assess antibody adherence inhibition activities as previously described (21, 23, 24). Briefly, overnight cultures of 3030-2 and 8516 bacteria were harvested by centrifugation (3,000 × g for 5 min) and resuspended in 5 ml PBS. Bacterial suspensions (100 μl; 1.5 × 10⁶ CFU) were mixed with 30 μl heat-inactivated mouse sera from the immunized groups (with or without dmLT adjuvant) or the control group on a rotating shaker (50 rpm) at room temperature for 30 min. Each mixture was then added to IPEC-J2 cells (1.5 × 10⁵) and incubated in a CO₂ incubator at 37°C for 1 h. After 3 washes with PBS to remove nonadherent bacteria, cells were incubated with sterile 0.5% Triton X-100 (500 μl per well) at room temperature for 30 min. Dislodged IPEC-J2 cells (with adherent bacteria) were collected by centrifugation, suspended in PBS, serially diluted, and plated on LB agar plates. After overnight incubation at 37°C, the numbers of bacterial colonies (CFU) on the plates were counted.

Mouse serum antibody neutralization assays against CT and STa enterotoxicity.

Human colon carcinoma cell line T-84 (ATCC; CCL-248) and direct cAMP and cGMP EIA kits (Enzo Life Sciences, Inc., Farmingdale, NY) were used to measure mouse serum antibody neutralization activity against CT and STa enterotoxicity, as described previously (12, 16). Briefly, T-84 cells, after reaching 95% confluence in Dulbecco’s modified Eagle’s medium (DMEM) plus Ham’s F-12 medium (DMEM/F-12; 1:1 [Invitrogen, Carlsbad, CA]) with 5% fetal bovine serum (FBS), were gently rinsed with PBS and then incubated with 700 μl cell medium (without FBS) with 1 mM IBMX (3-isobutyl-1-methylxanthine) at 37°C in 5% CO₂ for 25 min. Ten nanograms of CT (a homologue of LT; Sigma, St. Louis, MO) or 2 ng STa premixed with 30 μl heat-inactivated mouse serum samples from the immunized groups (with or without dmLT adjuvant) or control group at room temperature for 30 min on a rotary shaker (25 rpm) was brought to 300 μl with DMEM/F-12 medium and transferred to T-84 cells in each culture plate well. After incubation at 37°C in 5% CO₂ for 3 h (CT) or 1 h (STa), T-84 cells were washed with sterile PBS and dislodged and lysed with 0.1 M HCl with 0.5% Triton X-100. Cell lysates were collected by centrifugation (1,000 × g for 10 min) and examined for intracellular cGMP or cAMP levels using an EIA kit (Enzo Life Science) by following the manufacturer’s protocol.

Mouse serum antibody neutralization assays against STb and Stx2e cytotoxicity.

Vero cells (CCL-81; ATCC) and bacterial filtrates from STb⁺ strain 8020 and Stx2e⁺ strain 9168 were used to assess mouse serum antibody neutralization activities against STb and Stx2e toxins with a Vero cell cytotoxicity assay as described previously (20). No purified STb or Stx2e toxin was available in the laboratory. Vero cells at 95% confluence in Eagle’s minimum essential medium (EMEM) with 5% fetal bovine serum (FBS) were rinsed with sterile PBS and cultured in cell medium (without FBS). For the STb cytotoxicity assay, 300 μl culture filtrates of STb⁺ strain 8020 (CD₅₀ to Vero cells) premixed with 150, 25, or 10 μl heat-inactivated mouse serum samples from the immunized groups (with or without dmLT) or control groups were added to Vero cells (in a final volume of 700 μl) and incubated at 37°C in 5% CO₂ for 1 h. For the Stx2e cytotoxicity assay, 100 μl bacterial filtrates of Stx2e⁺ strain 9168 (CD₅₀) premixed with 18.8, 6.3, or 3 μl heat-inactivated mouse serum samples from the immunized groups (with or without dmLT) or control groups were added to Vero cells (in a final volume of 1,000 μl) and incubated at 37°C in 5% CO₂ for 3 days. Cells were examined with a microscope. Each cytotoxicity assay using the same STb⁺ or Stx2e⁺ bacterial culture filtrates (kept in a freezer) was conducted two times.

Statistical analyses.

Data were analyzed using GraphPad Prism version 7.0.0 (GraphPad Software, San Diego, CA). One-way analysis of variance (ANOVA) was used to examine differences at antifimbria and antitoxin antibody titers and antibody neutralization activities between the immunization group and the control group. A P value of <0.05 was considered statistically significant. Results are presented as means ± standard deviation (SD). All experiments were repeated two times using duplicate samples.