key: cord-0887005-u3ivh4o5
authors: Rausch, Dana; Ruan, Xiaosai; Nandre, Rahul; Duan, Qiangde; Hashish, Emad; Casey, Thomas A.; Zhang, Weiping
title: Antibodies derived from a toxoid MEFA (multiepitope fusion antigen) show neutralizing activities against heat-labile toxin (LT), heat-stable toxins (STa, STb), and Shiga toxin 2e (Stx2e) of porcine enterotoxigenic Escherichia coli (ETEC)
date: 2017-04-30
journal: Veterinary Microbiology
DOI: 10.1016/j.vetmic.2016.02.002
sha: 191350faa97b3279eb49a1b4135ed0a6a6d8985b
doc_id: 887005
cord_uid: u3ivh4o5

Abstract Enterotoxigenic Escherichia coli (ETEC) strains are the main cause of diarrhea in pigs. Pig diarrhea especially post-weaning diarrhea remains one of the most important swine diseases. ETEC bacterial fimbriae including K88, F18, 987P, K99 and F41 promote bacterial attachment to intestinal epithelial cells and facilitate ETEC colonization in pig small intestine. ETEC enterotoxins including heat-labile toxin (LT) and heat-stable toxins type Ia (porcine-type STa) and type II (STb) stimulate fluid hyper-secretion, leading to watery diarrhea. Blocking bacteria colonization and/or neutralizing enterotoxicity of ETEC toxins are considered effective prevention against ETEC diarrhea. In this study, we applied the MEFA (multiepitope fusion antigen) strategy to create toxoid MEFAs that carried antigenic elements of ETEC toxins, and examined for broad antitoxin immunogenicity in a murine model. By embedding STa toxoid STaP12F (NTFYCCELCCNFACAGCY), a STb epitope (KKDLCEHY), and an epitope of Stx2e A subunit (QSYVSSLN) into the A1 peptide of a monomeric LT toxoid (LTR192G), two toxoid MEFAs, ‘LTR192G-STb-Stx2e-STaP12F’ and ‘LTR192G-STb-Stx2e-3xSTaP12F’ which carried three copies of STaP12F, were constructed. Mice intraperitoneally immunized with each toxoid MEFA developed IgG antibodies to all four toxins. Induced antibodies showed in vitro neutralizing activities against LT, STa, STb and Stx2e toxins. Moreover, suckling piglets born by a gilt immunized with ‘LTR192G-STb-Stx2e-3xSTaP12F’ were protected when challenged with ETEC strains, whereas piglets born by a control gilt developed diarrhea. Results from this study showed that the toxoid MEFA induced broadly antitoxin antibodies, and suggested potential application of the toxoid MEFA for developing a broad-spectrum vaccine against ETEC diarrhea in pigs.

Young piglets commonly develop diarrhea, clinical conditions known as neonatal diarrhea and post-weaning diarrhea. Pig neonatal diarrhea and post-weaning diarrhea are caused by pathogenic bacteria and viruses including diarrheal Escherichia coli, coronaviruses (transmissible gastroenteritis virus -TGE, and porcine epidemic diarrhea virus -PEDV) and rotaviruses, and continue to be the most important swine diseases (Fairbrother et al., 2005; Harvey et al., 2005; USDA, 2002) . Diarrhea results in weight loss, slow growth and acute death, causing substantial economic losses to swine producers in the U.S. and other countries (Haesebrouck et al., 2004; Nagy and Fekete, 1999; Verdonck et al., 2002; Vu-Khac et al., 2007) . While neonatal diarrhea can be effectively prevented by passive protection of maternal antibodies, through immunization of pregnant sows; there are no effective prevention strategies against post-weaning diarrhea in weaned pigs. Vaccination is considered the most practical and likely the most effective approach to prevent pig diarrhea. Unfortunately, there are no broadly effective vaccines available to protect young pigs against diarrhea particularly post-weaning diarrhea (Melkebeek et al., 2013; Ruan et al., 2011) .

E. coli strains have a central role in the etiology of pig diarrhea (Hampson, 1994) . Among the types of E. coli causing diarrhea in pigs (Nataro and Kaper, 1998) , enterotoxigenic E. coli (ETEC, viz. a group of E. coli strains producing enterotoxins) is by far the most common and important. These ETEC strains produce fimbrial adhesins and enterotoxins. Fimbrial adhesins mediate ETEC bacterial attachment to specific receptors at pig intestinal epithelial cells and facilitate colonization in the pig small intestine (Smith and Linggood, 1971) , playing an essential role in initiating the disease. But it is the enterotoxins that disrupt fluid homeostasis in pig small intestinal epithelial cells to cause fluid and electrolyte hyper-secretion, leading to diarrhea (Nataro and Kaper, 1998) . Fluid hyper-secretion results in the loss of some of the mucin layer and disruption of tight junctions of microvilli that further enhance bacterial colonization and worsen diarrheal disease (Berberov et al., 2004; Glenn et al., 2009; Zhang et al., 2006) . Therefore, fimbrial adhesins and enterotoxins are the key virulence determinants of ETEC diarrhea, and have been long targeted for prevention strategy development.

Fimbrial adhesins expressed by ETEC strains causing diarrhea in pigs are K88 (F4), F18, K99 (F5), 987P (F6), and F41 (F7) (Awad-Masalmeh et al., 1982; Casey et al., 1992; Frydendahl, 2002; Moon, 1990; Moseley et al., 1986; Nagy et al., 1977; Zhang et al., 2007) . Toxins produced by porcine ETEC strains are porcinetype heat-labile toxin (pLT; which is homologous to but differs from the LT of ETEC causing human diarrhea-hLT), heat-stable toxin type Ia (pSTa or porcine-type STa), heat-stable toxin type II (STb), Shiga toxin type 2e (Stx2e), and enteroaggregative heatstable toxin type 1 (EAST1) (Frydendahl, 2002; Lee et al., 1983; Moon et al., 1980; Nakazawa et al., 1987; Osek, 1999; Zhang et al., 2007) . Clinical observations and epidemiological studies revealed that a typical ETEC strain expresses one and occasionally two fimbrial adhesins and one, two or more toxins (Francis, 2002; Frydendahl, 2002; Zhang et al., 2007) . Laboratory experimental studies demonstrated that an ETEC strain expressing one fimbrial adhesin and LT, STb, or STa enterotoxin causes diarrhea in young pigs (Berberov et al., 2004; Erume et al., 2008; Zhang et al., 2006; Zhang et al., 2008) . Unlike LT and STa or STb, Stx2e is a member of the Shiga toxin family, and itself is thought to be primarily associated with Edema disease in young pigs (Bertschiner and Gyle, 1994) . But as Stx2e is frequently found in ETEC strains expressing LT and/or ST toxins (Zajacova et al., 2012; Zhang et al., 2007) , it was also targeted for ETEC diarrhea prevention. In contrast, E. coli strains expressing fimbriae and EAST1 are found not associated with diarrhea in young pigs (Ruan et al., 2012; Zajacova et al., 2013) . It is thought that a vaccine blocking attachment from all ETEC fimbrial adhesins to host receptors and/ or eliminating enterotoxicity of ETEC toxins to host epithelial cells would be able to effectively protect against ETEC diarrhea in pigs and humans (Boedeker, 2005; Walker, 2005; Zhang and Sack, 2012) .

However, developing effective vaccines against ETEC continues to be difficult because ETEC strains are divergent. Different ETEC bacteria express immunologically heterogeneous fimbrial adhesins and distinctive enterotoxins. To overcome this challenge, innovative antigen preparation approaches are needed. The recently developed MEFA (multiepitope fusion antigen) strategy allowed us to include antigenic elements (epitopes) from multiple human ETEC virulence factors into a single recombinant protein to induce broadly protective antibody responses (Ruan et al., 2015; Ruan et al., 2014a) . We also found recently that peptides (or toxoids) derived from mutated LT and STa toxins are safe antigens and LT-STa genetic fusions induce protective antibodies against both LT and ST toxins Ruan et al., 2014b; Zhang et al., 2010) . By applying the MEFA, toxoid, and genetic fusion approaches, we should be able to include antigenic elements from all porcine ETEC toxins into a single antigen and to develop a safe and broadly effective antitoxin vaccine.

In this study, we applied the toxoid antigens and the MEFA strategy to create LT R192G -STb-Stx2e-STa P12F toxoid MEFAs, examined the toxoid MEFA for broad antitoxin antigenicity, and explored potential application of the toxoid MEFA in development of a broadly effective antitoxin vaccine against ETEC-associated diarrhea in pigs.

The E. coli strains and plasmids used in this study are listed in Table 1 . Genomic DNA of porcine E. coli field strains 3030-2 (K88/ LT/STb/STa) (Zhang et al., 2006) and 9168 (F18/Stx2e) (Zhang et al., 2007) , and STb recombinant strain 8020 (Zhang et al., 2006) were used for PCR amplification of the eltAB genes (LT), the estA gene (STa), the Stx2e A subunit gene and the estB gene (STb), respectively. LT mutant strain 8221 (Zhang et al., 2010) and STa mutant strain 8415 (Zhang et al., 2010) were used for LT R192G and STa P12F gene amplification. Strains 9168 and 8020 were also used in Vero cell test to detect antibody neutralizing activity against Stx2e toxin and STb toxin. Recombinant ETEC strains, 8819 (987P/LT), 8816 (987P/STb) and 8823 (987P/STa) were used in piglet challenge studies. Vector pET28a (Novagen, Madison, WI) was used to clone and to express the toxoid MEFA genes, and vector pMAL-p5X (New England Biolabs, Ipswich, MA) was used to clone the estB gene and the Stx2e A subunit gene for expression of MBP (maltose binding protein)-STb and MBP-Stx2e fusion proteins. E. coli strains BL21 (GE Healthcare, Piscataway, NJ) and DH5a (Promega, Madison, WI) were used to express toxoid MEFA and MBP fusion proteins. Recombinant E. coli strains were cultured in Lysogeny broth (LB) supplemented with kanamycin (30 mg/ml) or ampicillin (100 mg/ml).

B-cell epitopes from the LT A1 peptide, STb toxin, and the Stx2e A subunit were predicted with web-based software (Odorico and Pellequer, 2003; Saha and Raghava, 2007) . Full-length STa toxoid STa P12F gene was first embedded into LT A1 of a modified LT gene coding a monomeric LT R192G (one A subunit and one B subunit into a single peptide); nucleotides coding the STb epitope and the Stx2e A subunit epitope were sequentially embedded into the LT R192G -STa P12F chimeric gene by using SOE (splice overlapping extension) PCR as we described previously Ruan et al., 2014b; Zhang et al., 2010) .

To embed nucleotides coding the STa P12F mature peptide into the LT A1 of LT R192G , two PCR products were generated and overlapped. The first PCR product was amplified with primers LT 192 NheI-F3 (5-GTTTGCTAGCAATGGCGACAAATTATAC-0 3; NheI site underlined) and LT 192 -STa P12F -R (5 0 -GGCAAAATTACAACAAAGTT-CACAGCAGTAAAATGTGTTGTTTTCATCAATCACACCAAAATTAACACGA TACCA-0 3; nucleotides of STa P12F in italic, with mutated nucleotides in shade; nucleotides of LT A1 underlined). The second PCR product was amplified with primers LT 192 -STa P12F -F (5 0 -CTTTGTTGTAA TTTTGCCTGTGCTGGATGTTATATAGCTCCGGCAGAGGATGGTTACAGA-0 3; nucleotides of STa P12F in italic, with mutated nucleotides in shade; nucleotides of LT A1 underlined) and LT 192 BamHI-R1 (5 0 -GCGTGGATCCCTACTAGTTTTCCATACT-0 3; BamH1 site underlined). LT R192G plasmid p8221 was used as DNA templates for both PCRs. Overlapping two PCR products (through the STa complementary nucleotides of primers) generated a LT R192G -STa P12F chimeric gene. Digested with NheI and BamH1 (Biolabs), this LT R192G -STa P12F toxoid chimeric gene was cloned into pET28a. Resultant plasmids were introduced into DH5a competent cells for a LT R192G -STa P12F recombinant strain.

Similar to the insertion of STa P12F into the LT A1 of LT R192G , nucleotides coding the STb epitope and the Stx2e A subunit epitope were sequentially inserted into the A1 fragment of the LT R192G -STa P12F chimeric gene. To insert the STb epitope to create chimeric gene LT 192 -STb-STa P12F , we overlapped two PCR products: one was amplified with primers T7-F (5 0 -TAATACGACTCACTATAGGG-0 3) and

STb-LT A Chim-R (5 0 -ATGTTCACACA-GATCTTTTTTGCCGGTTTGTGTTCCTCTCGCGTG-0 3; nucleotides of STb epitope in italic, and nucleotides of LT A1 underlined), and the other was amplified with primers LT A -STbChim-F (5 0 -GATCTGTGTGAACATTATGTTTCCACTTCTCTTAGTTTGAGAAGT-0 3; nucleotides of STb epitope in italic, and nucleotides of LT A1 underlined) and T7-R (5 0 -TGCTAGTTATTGGTCAGGGGT-0 3), with plasmid pLT R192G -STa P12F as the DNA template.

Insertion of nucleotides coding the Stx2e A subunit epitope into chimeric gene LT 192 -STb-STa P12F was completed by overlapping another two PCR products. One was amplified with T7-F and Stx2e-LT A -R (5 0 -CGAAGATACATAACTTTGTTGTATATACTGTCCTGCTAAGT-GAGC-0 3; nucleotides of Stx2e A subunit epitope in italic, and nucleotides of LT A1 underlined), and the other was amplified with primers LT A -Stx2e-F (5 0 -AGTTATGTATCTTCGTTAAATTATATCGTTA-TAGCAAATATGTTT-0 3; nucleotides of Stx2e A subunit epitope in italic, and nucleotides of LT A1 underlined) and T7-R, with plasmid pLT R192G -STb-STa P12F as the DNA template. The overlapped PCR product was amplified with primers T7-F and T7-R, digested with NheI and BamH1, and cloned into pET28a.

To enhance anti-STa antigenicity, two additional copies of STa P12F gene were fused at the 5 0 end and the 3 0 end of the LT R192G -STb-Stx2e-STa P12F gene for a toxoid MEFA which carries three copies of STa toxoid STa P12F (LT R192G -STb-Stx2e-3xSTa P12F ).

Expression of toxoid MEFA proteins was examined in a standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Toxoid MEFA recombinant strains were cultured at 37 C in 200 ml Lysogeny broth (LB) supplemented with kanamycin (30 mg/ ml). Bacteria in the overnight grown culture, after OD 600 reached 0.6, were induced with 0.1 mM isopropyl-1-thio-b-D-galactoside (IPTG) and incubated for 4 more hours. Induced bacteria were collected and used for total protein extraction in B-PER (bacterial protein extraction reagent in phosphate buffer; Pierce, Rockford, IL). B-PER extracted proteins (with a majority of proteins as inclusion bodies; in denaturing buffer) were used for further extraction of 6 Â His-tagged toxoid MEFA proteins to a purity of greater than 90% with Ni-nitrolotriacetic acid agarose by following the manufacturer's protocol (QIAGEN, Valencia, CA). The 6 Â His tagged toxoid MEFA proteins were refolded with a Novagen protein refolding kit (Novagen).

The 6xHis-tagged protein (10 mg) of either toxoid MEFA recombinant strain was examined using 12% SDS-PAGE gels with rabbit anti-CT (1:3,000; Sigma, St. Louis, MO), anti-STa (1:5,000; protein A-purified, gift from Dr. DC Robertson, Kansas State Univ.), anti-STb (1:1,000; National Animal Disease Center, Ames, Iowa), and anti-Stx2e serum (1:3,000; National Animal Disease Center). IRDye-labeled goat anti-rabbit IgG (1:5,000; LI-COR, Lincoln, NE) and LI-COR Odyssey premium infrared gel imaging system (LI-COR) were used to detect the toxoid MEFA proteins. Protein purity was assessed using standard Coomassie blue staining as previously described (Ruan et al., 2014a) .

Ten 8-week-old female BALB/c mice (Charles River Laboratories International, Inc., Wilmington, MA) in a group were each immunized intraperitoneally with 200 mg of 6 Â His-tagged toxoid Table 1 Escherichia coli strains and plasmid used in the study.

Relevant properties Sources

fhuA2, D(argF-lacZ), U169, phoA, glnV44, w80, D(lacZ)M15, gyrA96, recA1, relA1, endA1, thi-1,hsdR17 (boosters) served as the control. Serum samples collected before the primary and 10 days after the final booster were stored at À80 C until use.

On day 38, all mice were anesthetized with CO 2 and exsanguinated. The mouse study complied with the Animal Welfare Act according to National Research Council guidelines (National Research Council, 1996) , and was approved by the South Dakota State University Institutional Animal Care and Use Committee and was supervised by the South Dakota state veterinarian.

Serum of each immunized or control mouse was examined for anti-LT, anti-STa, anti-STb and anti-Stx2e IgG antibodies in ELISAs. To titrate anti-LT and anti-STa IgG antibodies, we coated 96-well microtiter plates with cholera toxin (CT, an LT homologue which has been commonly used as the coating antigen to titrate anti-LT antibodies, Sigma; 100 ng per well of Immulon 2HB plates) or STaovalbumin conjugates (10 ng per well of Costar plates) as we described previously Ruan et al., 2014b; Zhang et al., 2010) .

To titrate anti-STb and anti-Stx2e IgG antibodies, we prepared MBP (maltose binding protein)-STb and MBP-Stx2eA fusion proteins and used them as coating antigens. The STb gene was PCR amplified with primers STbNcolI-F (5 0 -GACCTTTCATC-CATGGGCTCTACACAATCAAATAAAAAAGAT-0 3; Nco1 site underlined) and STbBamH1-R (5 0 -GACTTTAGAGGATCCTTAGCATCCTTTTGCTGCAACCAT-0 3; BamH1 site underlined), with STb plasmid p8020 (pRAS1) as templates. The Stx2e A subunit gene was amplified with primers Stx2eNco1-F (5 0 -GACCTTTCATCCATGGGCTTACTGGGTTTTTCTTC GGTATCC-0 3; Nco1 site underlined) and Stx2eBamH1-R (5 0 -GACTTTAGAGGATCCCACGTCTTCCGGCGTCATCGTATA-0 3; BamH1 site underlined) with total DNA of field strain 9168 (F18/ Stx2e) as templates. PCR amplified STb gene and Stx2e A subunit gene were digested with Nco1 and BamH1, and ligated into pMAL-p5x (BioLabs) for expression of MBP-STb and MBP-Stx2eA fusion proteins. Recombinant fusion proteins MBP-STb and MBP-Stx2eA expressed in E. coli DH5a were extracted with the pMAL TM protein fusion and purification system by following the manufacturer's protocol (BioLabs). Extracted MBP-STb and MBP-Stx2eA fusion proteins were refolded with a Novagen protein refolding kit (Novagen), verified in Western blot with anti-STb and anti-Stx2e antiserum respectively, and were used to coat Immulon 2HB plates (100 ng per well) to titrate anti-STb and anti-Stx2e IgG antibodies in ELISAs.

Pooled serum samples from the mice immunized with MEFA LT R192G -STb-Stx2e-3xSTa P12F and the control mice were examined for in vitro antibody neutralization activities against LT, STa, STb, and Stx2e. EIA cAMP and cGMP kits (Assay Design, Ann Arbor, MI) and T-84 cells were used to measure neutralizing activity of mouse serum antibodies against LT and STa toxins, as we described previously Ruan et al., 2014b; Zhang et al., 2010) .

Briefly, the serum sample (30 ml) pooled from the immunized group or the control group was incubated with 2 ng STa toxin (gift from Dr. DC Robertson, Kansas State Univ.; diluted in 150 ml Dulbecco modified Eagle medium [DMEM]/F-12 medium) or 10 ng CT (Sigma; in 150 ml DMEM/F-12) for 1 hour at room temperature. Incubated serum/toxin mixture was brought to 300 ml with DMEM/F-12 and added to a confluent monolayer of T-84 cells (10 5 cells per well, in 700 ml DMEM/F-12) pre-treated with 1 mM IBMX (3-isobutyl-1-methylaxanthine; Sigma). After incubation of 1 h (STa for cGMP) or 3 h (CT for cAMP) at 37 C in a CO 2 incubator, T-84 cells were gently washed and then lysed. Cell lysates were collected and were measured for intracellular cAMP or cGMP levels (pmole/ml) using a cAMP or cGMP ELISA kit by following manufacturer's protocol (Assay Design).

For antibody neutralization activity against STb toxicity, bacterial culture filtrates of strain 8020 (STb + ) and Vero cells (ATCC; CCL-81) were examined for cytotoxic activity. Vero cells were seeded in a 24-well plate and cultured to reach full confluence. The volume of 8020 filtrates causing over CD50 of Vero cells was pre-determined and used for the in vitro neutralization assay. Mouse serum sample, at 150 ml, 100 ml, 50 ml or 25 ml, pooled from the immunized mice or the control mice was used to be pre-mixed with 300 ml 8020 culture filtrates. Each serum/filtrates mixture was added to Vero cells (in 700 ml Eagle's Minimum Essential Medium). After incubation in a 37 C CO 2 incubator for 1 h, cells were examined microscopically. The highest dilution that prevented bacteria filtrates from causing CD50 of Vero cells was considered antibody neutralization titer.

To examine antibody neutralization activity against Stx2e toxin, Vero cell test was carried out based on published protocols (Matsuura et al., 2009; Smith and Scotland, 1993) , with modification. A volume of filtrates of Stx2e strain 9168 that caused CD50 of Vero cells was pre-determined. That volume (100 ml) of filtrates of 9168 overnight grown culture was mixed with binary diluted mouse serum sample, 50 ml, 25 ml, 12.5 ml, 6.3 ml, 3 ml, or 1.5 ml, pooled from the immunized or the control group. Each mixture was added to Vero cells (confluent monolayer) in a well of a 24well tissue culture plate (in a final volume of 1 ml with culture medium) and incubated in a CO 2 at 37 C for 3 days. Cells were examined microscopically daily for cell death (round up) and detachment from wells of the tissue culture plates. Vero cells cultured in cell medium alone or with 100 ml bacterial filtrates were used as references.

Gilts that had no ETEC diarrhea record and had not been vaccinated were used for this study. Gilt pre-screened without pre-existing antibodies to ETEC toxins were raised in the university swine unit. A pregnant gilt was intramuscularly immunized with 500 mg recombinant 'LT R192G -STb-Stx2e-3xSTa P12F ' (in 500 ml PBS) and 5 mg double mutant LT adjuvant (dmLT, in 50 ml PBS; provided by Walter Reed Army Institute of Research, Silver Spring, MD) 8 weeks before farrowing, and received a booster 4 weeks later. Another gilt IM injected with 500 ml PBS was used as the control. After 24 h suckling, piglets born by the immunized mother and the control mother were orally inoculated with 8819 (2.5 Â 10 9 CFUs), 8816 (2.5 Â 10 9 CFUs) and 8823 (2.5 Â 10 9 CFUs). Challenged piglets were examined every 3-4 h during 24 h post-inoculation. No piglets were challenged with a Stx2e strain. Pig immunization and challenge studies were approved and supervised by Kansas State University IACUC.

Data generated from this study were analyzed using SAS for Windows, version 8 (SAS Institute, Cary, NC). Results were expressed as means and standard deviations. Nonparametric Mood's Median Test or Kruskal-Wallis Median Test was carried out using SigmaXL (Kitchener, ON, Canada) at 2-sided and 95% confidence to assess differences at results of antibody titration (OD readings or the serum dilution that gave an OD >0.3 above the background) and antibody neutralization (cAMP or cGMP levels, pmole/ml) studies between the immunized group and the control group. Calculated p values of <0.05 indicated that differences were significant, when treatments were compared with a two-tailed distribution and two-sample equal or unequal variance.

3.1. The constructed toxoid MEFA carried epitopes of STb and Stx2e A subunit, and STa toxoid STa P12F By substituting an in silico predicted surface-exposed but less antigenic epitope of the monomeric LT R192G (138-154 amino acids of the LT A1 peptide) with the full-length mature peptide of STa P12F toxoid, a 'LT R192G -STa P12F ' toxoid chimeric gene was generated. Toxoid STa P12F was reported to induce neutralizing anti-STa antibodies in rabbits after being genetically fused at the Cterminus of monomeric LT R192G peptide (Zhang et al., 2010) . This 'LT R192G -STa P12F ' was served as the template for subsequent insertions of a STb epitope and a Stx2e epitope. Peptide 'KKDLCEHY' was in silico predicted as the only epitope from the poorly immunogenic STb toxin, and 'QSYVSSLN' was among the most antigenic epitopes from the Stx2e A subunit. Replacing nucleotides coding the 52-59 amino acids of the LT A1 in the 'LT R192G -STa P12F ' chimeric gene with nucleotides coding the STb 'KKDLCEHY' epitope resulted in 'LT R192G -STb-STa P12F ' gene. Substituting nucleotides coding the 77-83 amino acids of the LT A1 of 'LT R192G -STb-STa P12F ' with nucleotides coding the Stx2e A subunit epitope 'QSYVSSLN' yielded the 'LT R192G -STb-Stx2e-STa P12F ' gene ( Fig. 1A) . Accordingly, recombinant strains, 8778 (LT R192G -STa P12F ), 9137 (LT R192G -STb-STa P12F ), and 9161 (LT R192G -STb-Stx2e-STa P12F ) were generated (Table 1) .

Fusing two additional STa P12F toxoids at the N-terminus and the C-terminus of the 'LT R192G -STb-Stx2e-STa P12F ' created the second toxoid MEFA, recombinant strain 9403 (LT R192G -STb-Stx2e-3xSTa P12F ) ( Table 1 ). DNA sequencing revealed that nucleotide fragments coding the STa P12F toxoid, and epitopes of STb and Stx2e A subunit stayed in a correct reading frame.

Expression of the 'LT R192G -STb-Stx2e-STa P12F ' or 'LT R192G -STb-Stx2e-3xSTa P12F ' toxoid MEFA protein was verified in Western blot using anti-CT, anti-STa, anti-STb, and anti-Stx2e antiserum. A protein of approximately 37-40 kDa, an expected molecule weight of the denatured his-tagged toxoid MEFA 'LT R192G -STb-Stx2e- STa P12F ', was detected from strain 9161 by anti-CT, anti-STa, anti-STb, and anti-StxII antiserum, respectively (Fig. 1B) . A protein of approximately 40 kDa was detected from strain 9403 (LT R192G -STb-Stx2e-3xSTa P12F ) using the anti-STa and anti-LT antibodies. No proteins of the similar sizes were detected by the same anti-serum from the host E. coli BL21 strain.

Both toxoid MEFAs were unable to stimulate cAMP or cGMP in T-84 cells, indicating a lack of LT or STa enterotoxicity (data not shown). Mice immunized with toxoid MEFA 'LT R192G -STb-Stx2e-STa P12F ' developed strong anti-LT, anti-STb and anti-Stx2e IgG antibody responses (Fig. 2) . Anti-LT, anti-STb and anti-Stx2e IgG titers (in log 10 ) were 3.75 AE 0.34, 3.37 AE 0.42 and 3.6 AE 0.53, respectively in the immunized mice. These titers were significantly greater than those of the control mice (anti-LT, 0.1 AE 0.1, anti-STb, 0.42 AE 0.61, and anti-Stx2e, 1.26 AE 0.08; p = 0.000 from Mood's Median test).

Anti-STa IgG antibody response were detected at a very low titer (0.45 AE 0.21) in the serum of the mice immunized with 'LT R192G -STb-Stx2e-STa P12F ' (Fig. 2) , but still significantly differed from the response in the control mice (p = 0.003 of Mood's Median test). Mice immunized with toxoid MEFA 'LT R192G -STb-Stx2e-3xSTa P12F ', which included three copies of the STa P12F toxoid, had anti-STa IgG antibody titers (in log 10 ) detected at 0.87 AE 0.66 from serum samples (Fig. 3) .

The pooled serum sample of the immunized mice showed neutralizing activities against CT (Fig. 4) and STa (Fig. 5) . The cAMP level in T-84 cells incubated with 10 ng CT and 30 ml pooled serum of the immunized mice was 15.84 AE 3.40 pmole/ml; the cAMP level in the T-84 cells incubated with the same amount of CT but the pooled serum of the control mice was significantly greater, 41. 

The pooled serum sample of the immunized mice reduced STb cytotoxicity to Vero cells (Fig. 6) . Vero cells incubated with 300 ml 8020 filtrates alone (Fig. 6B) , or 300 ml filtrates and 150 ml pooled serum of the control mice (Fig. 6C) , were shown over 50% cell death or detachment. Cells incubated with 300 ml 8020 filtrates and 150 ml (1:7.7 dilution) pooled serum of the immunized mice remained normal (Fig. 6D ). When serum of the immunized mice was reduced to 25 ml (1:41 dilution), cells started showing 50% detachment (Fig. 6F) .

Stx2e toxicity Vero cell test showed 50% of the Vero cells became rounded and detached after incubation with 100 ml filtrates of Stx2e strain 9168 overnight grown culture (Fig. 7B) or in the addition of 50 ml pooled serum sample of the control mice (Fig. 7C ). But cells remained normal when incubated with 100 ml 9168 filtrates and 50 ml pooled serum sample (1:20 final dilution, in 1 ml cell culture medium) of the immunized mice (Fig. 7D) . Only when less than 12.5 ml pooled serum of the immunized mice was used (a final dilution of 1:80), over 50% of the cells become dead and detached (Fig. 7F) .

Immunized gilt and piglets born by the immunized gilt had anti-LT, anti-STb, anti-STa and antiStx2e antibodies detected Fig. 2 . Mouse serum anti-LT, anti-Stx2e, anti-STb and anti-STa IgG antibody titers (in log 10 ). Anti-LT, anti-Stx2e, anti-STb and anti-STa IgG antibodies in serum samples of the mice immunized with 'LT R192G -STb-Stx2e-STa P12F ' MEFA and the control mice (10 mice per group) were titrated in ELISAs using CT (100 ng per well of 2HB plates; coating antigens), MBP-Stx2eA fusion (100 ng per well of 2HB plates), MBP-STb fusion (100 ng per well of 2HB plates), and STa-ovalbumin conjugates (10 ng per well of Costar plates), respectively. HRP-conjugated goat anti-mouse IgG (1:5000 for anti-LT, anti-Stx2e and anti-STb; 1:3000 for anti-STa) was used the secondary antibodies. Antibody titers were calculated from the highest dilution of a serum sample that produced an ELISA OD of >0.3 (above the background), and expressed in a scale of log 10 . Each dot represented an IgG titer from a mouse, and the bar indicated the mean titer of the group. The p value indicated difference of the IgG titers between the immunization group and the control group.

( Table 2 ). After challenged with 8819 (LT), 8816 (STb) and 8823 (STa) recombinant ETEC strains together, all 6 piglets born by the control mother developed severe diarrhea and showed sign of dehydration, whereas 6 out 7 piglets born by the immunized mother remained healthy. Only one piglet born by the immunized mother showed mild diarrhea. No antibodies specific to LT, STa, STb or Stx2e were detected in the control gilt, piglets born by the control gilt, or serum samples of gilts prior to primary immunization.

Since ETEC strains expressing any one, two, or more than two toxins cause diarrhea in neonatal and post-weaning pigs, an effective ETEC antitoxin vaccine need to induce antibodies protecting against LT, STa and STb. Results from the present study indicated that a toxoid MEFA that carried antigenic elements or epitopes of four ETEC toxins induced anti-LT, anti-STa, anti-STb and anti-Stx2e antibodies. Moreover, antibodies derived from the immunized mice showed in vitro neutralization activities against toxicity of all four toxins, and antibodies derived from the immunized gilt showed protection against LT/STa/STb ETEC infection. These results suggest toxoid MEFA LT R192G -STb-Stx2e-3xSTa P12F can potentially be an antigen for developing a protective antitoxin vaccine against ETEC associated diarrhea in young pigs.

STa toxoid STa P12F when was fused to the C-terminus of the LT toxoid LT R192G induced strongly protective anti-STa IgG and IgA antibody responses in IM immunized rabbits (Zhang et al., 2010) . Data from the current study, however, showed that MEFA LT R192G -STb-Stx2e-STa P12F , which had the same STa toxoid but embedded inside the A1 peptide of the LT R192G , induced only mild IgG antibody response in the IP immunized mice. That suggested the STa toxoid may not be placed at an optimal position in this toxoid MEFA, negatively affecting its antigen presentation and thus anti-STa antigenicity. Knowing that additional copies of STa toxoid enhanced toxoid fusion anti-STa antigenicity , we constructed toxoid MEFA 'LT R192G -STb-Stx2e-3xSTa P12F ' to carry three copies of the STa P12F . Mice immunized with LT R192G -STb-Stx2e-3xSTa P12F developed similar levels of anti-LT, anti-STb and anti-Stx2e IgG antibody responses compared to the mice immunized with toxin MEFA 'LT R192G -STb-Stx2e-STa P12F ' (data not shown), but a greater anti-STa IgG antibody titer (Fig. 3) . Whether STa P12F is the optimal STa toxoid for toxoid MEFAs to induce protective anti-STa antibodies will need to be further examined. Human-type STa toxoid STa N12S is suggested an optimal STa toxoid for LT-STa toxoid fusions in inducing neutralizing anti-STa antibodies (Ruan et al., 2014b) . But the human-type STa and the porcine-type STa show antigenicity differences, whether the showed anti-LT antibody neutralization activity. Neutralizing antibodies prevent CT from stimulating intracellular cAMP in T-84 cells, resulting in a low cAMP level (pmole/ml). The pooled serum sample of the mice immunized with 'LT R19G -STb-Stx2e-3xSTa P12F ' or the control mice (30 ml in total) was incubated with 10 ng CT toxin (in 150 ml) for 1 h at room temperature, the serum/toxin mixture was brought to 300 ml and added to T-84 cells (in 700 ml cell culture medium). Intracellular cAMP levels in cells were measured after 3 h incubation. Ten ng CT without serum was used as a positive control, and medium (cell culture medium; without toxin or serum) was used as the background reference. Columns and bars indicated mean cAMP levels and standard deviations. The p value of Mood's Median test indicated difference of antibody neutralizing activity against CT between the immunization group and the control group. immunized with 'LT R19G -STb-Stx2e-3xSTa P12F ' or the control mice (30 ml in total) was incubated with 2 ng STa toxin (in 150 ml) for 1 h at room temperature, the serum/toxin mixture was brought to 300 ml and added to T-84 cells (in 700 ml cell culture medium). Intracellular cGMP levels in cells were measured after 1 h incubation. Two ng STa without serum was used as a positive control of enterotoxicity and medium (cell culture medium; without toxin or serum) was used as the background reference. Columns and bars indicated mean cGMP levels and standard deviations. The p value of Mood's Median test indicated difference of antibody neutralizing activity against STa between the immunization group and the control group. Fig. 3 . Anti-STa IgG titers (in log 10 ) in serum samples of mice immunized with toxoid MEFA 'LT R192G -STb-Stx2e-3xSTa P12F '. Anti-STa IgG antibodies in the serum samples of mice before the immunization and after the immunization (10 mice in the group) were titrated in ELISA using STa-ovalbumin conjugates (10 ng per well of Costar plates; coating antigen) and HRP-conjugated goat anti-mouse IgG (1:3000; the secondary antibodies). Each dot represented a mouse IgG titer, and the bar indicated the mean titer of the treatment.

analogue porcine-type STa N11S exhibits better antigenic property needs to be further examined. Nevertheless, antibodies induced by the LT R192G -STb-Stx2e-3xSTa P12F showed neutralizing activity against STa toxin, and protected suckling piglets against LT/STb/ STa ETEC infection.

LT R192G -STb-Stx2e-STa P12F and LT R192G -STb-Stx2e-3xSTa P12F generated in this study may be used to develop antitoxin subunit vaccines, as a step for further development of practical and economical vaccines to prevent diarrhea in pigs. Unlike live attenuated vaccines, subunit vaccines are not optimal for preventing post-weaning diarrhea in pigs because of a higher production cost and the demand for professional training and labor in parenteral immunization to large numbers of individual pigs. Helpfully, the monomeric 'LT R192G -STb-Stx2e-STa P12F ' MEFA had Fig. 6 . Mouse serum antibody neutralization activity against STb toxin. A: Normal Vero cells grown in cell culture medium. B: Vero cells (in 700 ml culture medium) incubated with 300 ml 8020 (STb) overnight grown culture filtrates, showing over 50% cell death and detachment. C: Vero cells incubated 300 ml 8020 (STb) filtrates pre-mixed with 150 ml pooled serum of the control mice. D: Vero cells incubated with 300 ml 8020 (STb) filtrates pre-mixed with 150 ml pooled serum of the mice immunized with 'LT R192G -STb-Stx2e-3xSTa P12F ' MEFA. E: Vero cells incubated with 300 ml 8020 (STb) filtrates premixed with 50 ml pooled serum of the immunized mice. F: Vero cells incubated with 300 ml 8020 (STb) filtrates and 25 ml pooled serum of the immunized mice. the STa toxoid, the STb epitope, and the Stx2e epitope embedded at the A1 peptide of LT R192G . This chimeric gene can be modified to express a LT-like holotoxin-structured toxoid MEFA protein. As we demonstrated previously (Ruan and Zhang, 2013) , by simply inserting back the nucleotides coding the native leading signal peptides of the LT A subunit gene (eltA) and the LT B subunit gene (eltB), and also the cistron gene structure between the two LT subunit genes, a monomeric LT gene can be reversed to a LT-like gene coding a LT-like holotoxin-structured protein. Since the native LT B gene expresses the LT B subunit and forms LT B pentamer independently, a modified 'LT R192G -STb-Stx2e-STa P12F ' fusion protein can form LT B pentamer and binds to GM1 receptors in pig small intestine. If expressed by a nonpathogenic E. coli strain, the modified LT-like-structured 'LT R192G -STb-Stx2e-STa P12F ' toxoid MEFA could be used to develop a live oral vaccine against porcine ETEC diarrhea. Even for the 'LT R192G -STb-Stx2e-3xSTa P12F ' toxoid MEFA, if all three copies of STa P12F or a different STa toxoid are embedded at the LT A1 peptide, for an example, at the N-and Cterminus of the A1 peptide, the monomeric 'LT R192G -STb-Stx2e-3xSTa P12F ' can also be converted to a holotoxin-structured protein and expressed by an E. coli or a Salmonella strain for a live ETEC vaccine candidate.

Double mutant LT, dmLT (LT R192G/L211A ) was used as adjuvant for pig immunization. This dmLT has been explored as a safe and effective mucosal adjuvant in human vaccine development. We recently showed that dmLTcan be an effective adjuvant in parenteral immunization (data not shown). Adjuvant dmLT included in current pig immunization also served as an antigen to supplement the toxoid MEFA in inducing protective anti-LTantibodies. Future studies with a different adjuvant can conclusively assess protection from anti-LT antibodies derived from this toxoid MEFA. Additionally, studies will be needed to optimize the dose of dmLT adjuvant and also the dose of toxoid MEFA immunogen.

Future studies to assess better the potency of this toxoid MEFA in ETEC vaccine development will be needed. The present study examined only antigen immunogenicity, in vitro antitoxin antibody neutralization and maternal antibody passive protection against ETEC infection. Unlike commercially available kits that were used to measure antibody neutralizing activities against LT or STa, we currently do not have standard assays to measure anti-STb and anti-Stx2e antibody neutralizing activities. In this study, we adopted the Vero cell cytotoxicity assay to assess antibody neutralization activities against STb and Stx2e. Future studies to use purified STb or Stx2e toxin instead of bacterial culture filtrates or to adopt a Shiga toxin quantitative microtiter assay (Gentry and Dalrymple, 1980 ) may measure better anti-Stx2e or anti-STb antibody neutralization activity. In addition, immunizing neonatal piglets and challenging weaned piglets with the LT+, STb+, and STa + ETEC strains separately will be needed to assess protective efficacy against each toxin, post-weaning diarrhea, and also perhaps Stx2e associated edema disease as described by Oanh et al. (2012) . Moreover, large scaled studies are required to evaluate efficacy of candidate vaccines derived from this toxoid MEFA. Results from this study, nevertheless, indicate the toxoid MEFAs can potentially serve as antigens for multivalent vaccines against ETEC associated diarrhea, and may suggest the MEFA strategy may be a useful platform for developing broadly protective vaccines against other heterogeneous pathogenic strains or isolates.

STa toxoid peptide, STb and Stx2e-A subunit epitopes of porcine enterotoxigenic E. coli (ETEC), the main cause of porcine neonatal diarrhea and post-weaning diarrhea, can be embedded into a LT toxoid monomer. Mouse serum antibodies derived against the toxoid MEFA showed neutralizing activities against all four toxins. Table 2 Anti-LT, -STb, -Stx2e and anti-STa IgG and IgA antibody titers (log 10 ) detected in the immunized or control sows and the piglets born by the immunized or control mothers. 

Pilus production, hemagglutination, and adhesion by porcine strains of enterotoxigenic Escherichia coli lacking K88, K99, and 987P antigens

Relative importance of heat-labile enterotoxin in the causation of severe diarrheal disease in the gnotobiotic piglet model by a strain of enterotoxigenic Escherichia coli that produces multiple enterotoxins

Oedema disease of pigs

Vaccines for enterotoxigenic Escherichia coli: current status

Pathogenicity of porcine enterotoxigenic Escherichia coli that do not express K88, K99, F41, or 987P adhesins

Comparison of the contributions of heat-labile enterotoxin and heatstable enterotoxin b to the virulence of enterotoxigenic Escherichia coli in F4ac receptor-positive young pigs

Escherichia coli in postweaning diarrhea in pigs: an update on bacterial types pathogenesis, and prevention strategies

Enterotoxigenic Escherichia coli infection in pigs and its diagnosis

Prevalence of serogroups and virulence genes in Escherichia coli associated with postweaning diarrhoea and edema disease in pigs and a comparison of diagnostic approaches

Quantitative microtiter cytotoxicity assay for Shigella toxin

Toxin-mediated effects on the innate mucosal defenses: implications for enteric vaccines

Efficacy of vaccines against bacterial diseases in swine: what can we expect?

Postweaning Escherichia coli diarrhoea in pigs

Use of competitive exclusion to control enterotoxigneic strains of Escherichia coli in weaned pigs

Characterization of the gene encoding heat-stable toxin II and preliminary molecular epidemiological studies of enterotoxigenic Escherichia coli heatstable toxin II producers

Heat-labile-and heat-stable-toxoid fusions (LT R192G -STa P13F of human enterotoxigenic Escherichia coli elicit neutralizing antitoxin antibodies

Novel subtilase cytotoxin produced by Shiga-toxigenic Escherichia coli induces apoptosis in vero cells via mitochondrial membrane damage

ETEC vaccination in pigs

Colonization factor antigens of enterotoxigenic Escherichia coli in animals

Prevalence of pilus antigens enterotoxin types, and enteropathogenicity among K88-negative enterotoxigenic Escherichia coli from neonatal pigs

Cloning of chromosomal DNA encoding the F41 adhesin of enterotoxigenic Escherichia coli and genetic homology between adhesins F41 and K88

Enterotoxigenic Escherichia coli (ETEC) in farm animals

Colonization of porcine intestine by enterotoxigenic Escherichia coli: selection of piliated forms in vivo adhesion of piliated forms to epithelial cells in vitro, and incidence of a pilus antigen among porcine enteropathogenic E. coli

Virulence factors in Escherichia coli isolated from piglets with neonatal and post-weaning diarrhea in Japan

Diarrheagenic Escherichia coli

Guide for the Care and Use of Laboratory Animals

BEPITOPE: predicting the location of continuous epitopes and patterns in proteins

Protection of piglets against Edema disease by maternal immunization with Stx2e toxoid

Prevalence of virulence factors of Escherichia coli strains isolated from diarrheic and healthy piglets after weaning

Oral immunization of a live attenuated Escherichia coli strain expressing a holotoxin-structured adhesin-toxoid fusion (1FaeG-FedF-LTA2:5LTB) protected young pigs against enterotoxigenic E. coli (ETEC) infection

A tripartite fusion, FaeG-FedF-LT(192) A2:B, of enterotoxigenic Escherichia coli (ETEC) elicits antibodies that neutralize cholera toxin, inhibit adherence of K88 (F4) and F18 fimbriae, and protect pigs against K88ac/heat-labile toxin infection

Escherichia coli expressing EAST1 toxin did not cause an increase of cAMP or cGMP levles in cells, and no diarrhea in 5-day old gnotobiotic pigs

Multiepitope fusion antigen induces broadly protective antibodies that prevent adherence of Escherichia coli strains expressing colonization factor antigen I (CFA/I) CFA/II, and CFA/IV

Characterization of heat-stable (STa) toxoids of enterotoxigenic Escherichia coli fused to a double mutant heat-labile toxin (dmLT) peptide in inducing neutralizing anti-STa antibodies

Genetic fusions of a CFA/I/II/IV MEFA (multiepitope fusion antigen) and a toxoid fusion of heat-stable toxin (STa) and heat-labile toxin (LT) of enterotoxigneic Escherichia coli retain broad anti-CFA and antitoxin antigenicity

Prediction methods for B-cell epitopes

Observations on the pathogenic properties of the K88, Hly and Ent plasmids of Escherichia coli with particular reference to porcine diarrhoea

isolation and identification methods for Escherichia coli O157 and other Vero cytotoxin producing strains

Part II: Reference for swine health and health management in the United States

Different kinetic of antibody responses following infection of newly weaned pigs with an F4 enterotoxigenic Escherichia coli strain or an F18 verotoxigenic Escherichia coli strain

Serotypes virulence genes, intimin types and PFGE profiles of Escherichia coli isolated from piglets with diarrhoea in Slovakia

Considerations for development of whole cell bacterial vaccines to prevent diarrheal diseases in children in developing countries

Detection of virulence factors of Escherichia coli focused on prevalence of EAST1 toxin in stool of diarrheic and non-diarrheic piglets and presence of adhesion involving virulence factors in astA positive strains

Experimental infection of gnotobiotic piglets with Escherichia coli strains positive for EAST1 and AIDA

Progressand hurdles inthe development ofvaccines against enterotoxigenic Escherichia coli in humans

Significance of heat-stable and heat-labile enterotoxins in porcine colibacillosis in an additive model for pathogenicity studies

Prevalence of virulence genes in Escherichia coli strains recently isolated from young pigs with diarrhea in the US

Escherichia coli constructs expressing human or porcine enterotoxins induce identical diarrheal diseases in a piglet infection model

Genetic fusions of heat-labile (LT) and heat-stable (ST) toxoids of porcine enterotoxigenic Escherichia coli elicit neutralizing anti-LT and anti-STa antibodies

Toxicity and immunogenicity of enterotoxigenic Escherichia coli heat-labile and heat-stable toxoid fusion 3xSTa A14Q -LT S63K/R192G/ L211A in a murine model

We thank Dr. DC Robertson for providing the anti-STa antiserum and the purified STa toxin. Financial support for this study was provided by the Agricultural Experiment Station of South Dakota State University and Kansas State University.

Pigs IM immunized with LT-STb-Stx2e-3xSTa P12F MEFA developed antibodies specific to each toxin, and maternal passive antibodies protected born piglets against infection of a mixture of a LT+, a STa+ and a STb+ strains. These results suggested that a toxoid MEFA may potentially serve as an antigen for vaccine development against porcine ETEC diarrhea.

The authors declaim no conflict of interests from this study.