Identifying Immuno-Dominant and Neutralizing Epitopes from K88 Identifying Immuno-Dominant and Neutralizing Epitopes from K88 Fimbriae of Enterotoxigenic Escherichia coli (ETEC) Fimbriae of Enterotoxigenic Escherichia coli (ETEC)

Summary Enterotoxigenic Escherichia coli (ETEC) bacteria are the primary cause of diarrheal disease, especially porcine post-weaning diarrhea (PWD). Post-weaning diarrhea is one of the most common diseases in piglets 3 to 10 days after weaning and causes the loss of millions of dollars annually to United States swine industry and other countries. 2 These ETEC bacteria produce two types of virulence factors: 1) fimbriae adhesins, which promote bacterial attachment and colonization in pig small intestine; and 2) enterotoxins that disrupt fluid homeostasis and cause fluid hype-secretion and watery diarrhea. The F4 (K88) is the most important fimbria in ETEC bacteria causing PWD. An effective vaccine against PWD would have to induce antibody responses against the K88 fimbri-ae. 3 In this study, we in silico identified epitopes from the K88 fimbriae of ETEC that are associated with pig neonatal diarrhea and PWD. We genetically fused each epitope to non-homologous human ETEC CFA/I adhesin subunit CfaB to present each K88 fimbrial major subunit FaeG epitope, and examined each fusion protein with anti-K88 antiserum to identify immunodominant epitopes. Furthermore, each epitope fusion was used to immunize mice. Mouse serum samples were titrated for IgG antibody response specific to K88 fimbriae. Mouse serum samples were further examined for antibody neutralization activity against adherence of K88 fimbriae using porcine cell lines IPEC-J2 and porcine ETEC wildtype strain 3030-2. To verify whether epitope conformation alteration could


Introduction
Enterotoxigenic Escherichia coli (ETEC) bacteria are the primary cause of swine postweaning diarrhea (PWD). Post-weaning diarrhea is one of the most common diseases in piglets 3 to 10 days after weaning and causes the loss of millions of dollars annually to United States swine industry and other countries. 4 These ETEC bacteria produce two types of virulence factors: 1) fimbriae adhesins, which promote bacterial attachment and colonization in pig small intestine; and 2) enterotoxins that disrupt fluid homeostasis and cause fluid hype-secretion and watery diarrhea. 2 The F4 (K88) is the most important fimbria in ETEC bacteria causing PWD. 2 An effective vaccine against PWD would have to induce antibody responses against the K88 fimbriae. 5 A recent study demonstrated that an adhesion-toxoid multiepitope fusion antigen (MEFA) induced protective antibodies against K88 and heat-labile toxin (LT) and suggested that this adhesin-toxoid MEFA can be an antigen for a vaccine against PWD. 6 However, this adhesin-toxoid MEFA does not carry antigens to induce antibodies against other enterotoxins, including heat-stable toxin II (STb). A more recent study showed that toxoid MEFA LT-STa-STb-Stx2e induced protective antibodies against all four ETEC toxins, but this toxoid MEFA did not carry ETEC fimbria antigens. 7 A broadly protective PWD vaccine likely needs to carry all the toxin antigens 4 Dubreuil, J., R. Isaacson, D. Schifferli. 2016 and the fimbria antigens. 8 To include antigenic elements from these ETEC toxins and fimbriae into a vaccine product, we applied structure-based vaccinology to develop an epitope-based PWD vaccine, by identifying neutralizing epitopes from each ETEC toxin and fimbria and then packing them into a single antigen component.

Bacteria Strain, Plasmid and Cell Line
The E. coli strains and plasmids used in this study are listed in Table 1. E. coli strain 9477 was used as DNA templates to PCR amplify the CfaB gene, 9 and E. coli strain 9503 for the expression of CfaB protein. Porcine ETEC field strain 3030-2 (K88/ LT/STb/STa) was utilized for FaeG subunit gene amplification. Vector pET28α was used to clone the CfaB-K88-epitope fusions and the FaeG subunit. Fusion proteins CfaB-K88-epitope and FaeG subunit protein were expressed by E. coli strains BL21 and DH5α in Lysogeny broth (LB) supplemented with kanamycin (30 μg/mL). Porcine intestinal cell line IPEC-J2 was used to examine antibody neutralizing activity against the adherence of K88 fimbriae.

Epitope Identification and Epitope Fusion Protein Preparation
Computer software was applied to in silico identify epitopes from FaeG major subunit of K88 fimbria. Each FaeG epitope was genetically fused to non-homologous human ETEC CFA/I adhesin subunit CfaB using splicing overlap extension (SOE) PCR with primers listed in Table 2. Digested with NheI and EagI, each epitope fusion PCR product was cloned into expression vector pET28a and expressed in E. coli BL21.
Epitope fusion recombinant strains were cultured in LB. Culture was grown overnight and transferred in 200 mL LB at a 1:50 ratio and continued to be cultured at 37°C. After OD 600 reached 0.5-0.6, bacteria were induced with 0.1 mM isopropyl-1-thio-βd-galactoside (IPTG) for 4 more hours. Induced bacteria were collected and used for total protein extraction using B-PER (bacterial protein extraction reagent in phosphate buffer).
Epitope fusion proteins were examined in a standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with anti-K88 antiserum. Epitope fusion proteins were also examined in ELISA with anti-K88 antiserum. The K88 fimbriae major subunit FaeG gene was amplified by PCR, cloned in pET28a, and expressed in E. coli BL21. Competitive ELISAs were carried out by coating K88 fimbriae (10 ng/well) to Immulon 2HB plates. Wells were incubated with anti-K88 mouse antiserum and each epitope fusion protein (800 ng/well) at 37°C for 1 h. The HRP-conjugated goatanti-mouse IgG (1:3000 dilution) was used as the secondary antibody. Optical densities (OD 650 ) were measured using an ELISA plate reader.

Mouse Immunization
To immunize each 8-week old BALB/c mouse (5 mice per group) subcutaneously, 40 μg of each epitope-CfaB fusion protein were used with 0.2 μg double mutant LT (dmLT) adjuvant. Immunized mice received two booster injections (the same dose of the primary) biweekly. Two weeks after the second booster injection, mice were euthanized. A group of five mice without immunization was used as the control.

Antibody Titration and Neutralization
Mouse serum samples collected two weeks after the final booster were utilized for antibody titration. Immulon 2HB plates coated with K88 fimbriae (100 ng/well) or FaeG recombinant protein (100 ng/well) were used to titrate epitope-specific antibodies. Mouse serum samples were further examined for antibody neutralization activity against adherence of K88 fimbriae using porcine cell line IPEC-J2. Western blot with FaeG recombinant protein or denatured K88 fimbriae were used to examine reaction with the serum of mice immunized with each epitope fusion.

Data Analysis
Two-way ANOVA was performed to analyze all ELISA data, and one-way ANOVA was used for antibody titration data and antibody adherence inhibition assay data. A P value of < 0.05 was considered as statistically significant. All experiments were repeated two times using duplicate samples.
To determine whether epitope fusion conformation alterations resulted in the loss of reaction with anti-K88 antiserum or if they fail in inducing antibody responses, ELISA and Western blot analyses were conducted using FaeG protein and denatured (boiled) K88 fimbriae (with FaeG subunits that were separated after boiling). Data showed that FaeG protein and the boiled K88 fimbriae, unlike the whole K88 fimbriae, were recognized by antibodies in the serum samples of all immunization groups ( Figure  4). These results indicate that epitopes LGRGGVTSADGEL, PRGSELSAGSA, and RENMEYTDGT may locate at the region connecting adjacent FaeG subunits for K88 fimbriae.