Novel fusion antigen displayed-bacterial ghosts vaccine candidate against infection of Escherichia coli O157:H7

Infection with Escherichia coli O157:H7 may develop into hemorrhagic colitis, or hemolytic uremic syndrome (HUS), which usually causes kidney failure or even death. The adhesion and toxins are the important virulent factors. In this study, a novel vaccine candidate rSOBGs was constructed based on the bacterial ghost (BG). rSOBGs maintained the integrity of cellular morphology and displayed the linear Stx2Am-Stx1B antigen on the surface of outer membrane. rSOBGs induced Stxs-specific IgA/IgG antibodies and stronger intimin-specific IgA/IgG antibodies effectively in sera in this study. In vivo, the rSOBGs provided the higher protection rate (52%) than native bacterial ghost-OBGs (12%) when challenged intragastricly with high dose (500 LD50) viable E. coli O157:H7. Meanwhile, the rSOBGs provided higher protection rate (73.33%) than OBGs when challenged with 2 LD50 even to 5 LD50 lysed E. coli O157:H7. In vitro, the rSOBGs-immunized sera possessed neutralizing activity to lysed pathogenic bacteria. Furthermore, the results of histopathology also displayed that the administration of rSOBGs have the ability to reduce or inhibit the adhesion lesions and toxins damages of organs. The novel vaccine candidate rSOBGs induced both anti-toxin and anti-adhesion immune protection, suggesting the possibility to prevent the infectious diseases caused by Escherichia coli O157:H7.

ghost (BG) is produced by the expression of PhiX174 lysis gene E, and results in cellular lysis and cytoplasmic loss 20 . BG maintains the cellular morphology and native surface antigenic structure 21 , and posses the adjuvant property 22 . The effects of BGs vaccines have been demonstrated in various pathogens, such as Vibrio cholerae 23 , Pasteurella haemolytica 24 and Helicobacter pylori 25 , and the BG has been developed to be the candidate vaccine against the infection of E. coli O157:H7 26,27 .
It is considered that co-infection of viable E. coli O157:H7 organism and toxins is general, and a novel vaccine which could provide the co-prevention of bacterial adhesion and toxic damage is imminent. In this study, the linear Stx2Am-Stx1B antigen was displayed on the surface of E. coli O157:H7 BGs based on a sandwich vector pSOmpA [28][29][30][31] which was constructed by the partial out membane protein A (OmpA) of Shigella dysenteriae and partial OmpA of E. coli to construct a novel candidate vaccine named rSOBGs. The immunogenicity, protection ability and immunologic mechanism against the challenge of E. coli O157:H7 were described subsequently.

Materials and Methods
Bacterial strains, plasmids, cell lines and media. Bacteria were grown in Luria-Bertani (LB) broth or agar (Oxoid) supplemented with 100 μ g/ml of ampicillin for selection of recombinant plasmid. LB broth or agar without ampicillin was used for culturing E. coli O157:H7. Hep-2 cells and Vero cells were conserved in the laboratory, and the cells were grown in Dulbecco's modified eagle medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS).

Ethics statement. The Ethics of Animal Experiments of Beijing Institute of Microbiology and
Epidemiology approved the study. The all experiments and methods were carried out in accordance with the approved relevant guidelines. BALB/c mice were maintained in the Animal Center under specific pathogen-free conditions in Beijing Institute of Microbiology and Epidemiology. BALB/c mice were fed with standard diet and water, maintained under the following conditions: 12 h light/12 h dark controlled lighting, 24 °C to 28 °C temperature, and 55% relative humidity. All animals were handled under the care and supervision of a veterinarian. The mice that were severely injured by E. coli O157:H7, were sacrificed by cervical dislocation at the end of experiment.
Construction of expression plasmid pOSAmB. The 1-233 amino acids of OmpA (Genbank: NC_007606) amplified from Shigella dysenteriaewas 51054 strain (primers: SOup/SOdown, table S1), and the 234-325 amino acids of OmpA amplified from E. coli O157:H7 EDL 933 strain (primers: EOup/ EOdown). Splicing overlap extension (SOE) PCR was introduced to generate SOmpA gene with primers SOdown/EOup (overlaid) and flanking primers SOup/EOdown. Subsequently, primers Enzup/Enzdown (overlaid) and flanking primers SOup/EOdown were used to introduce two restriction enzyme sites into SOmpA to generate ESOmpA gene. The ESOmpA gene was ligated into pMD18-T Simple vector to construct pESOmpA for sequencing, then plasmid was digested by EcoR I/Sac I, and the vector was harvested. SAmB gene 13 was amplified from pSAmB (primers Stxup/Stxdown), and restriction enzyme sites EcoR I/Sac I were introduced. Thereafter, the gene was cloned into pMD18-T Simple vector to generate pESAmB for sequencing, and the generated SAmB gene was ligated into digested pESOmpA vector, and the plasmid was digested by XmaI and Nco I to generate a gene named OSAmB. This OSAmB gene was ligated into the expression vector pGEX to generate the expression plasmid pOSAmB.
Preparation of rSOBGs. The correct pOSAmB plasmids were transformed into O157:H7 EDL933 as novel competent cells. The lysis plasmid pLysE (supp. Fig.1) was transformed into these novel competent cells to construct the vaccine strain of rSOBGs. Grown up to OD 600 0.3, the cultures of rSOBGs were induced by isopropyl β -D-thiogalactopyranoside (IPTG) at a final concentration of 1 mM at 28 °C. The induction of lysis was achieved by shifting the temperature from 28 °C to 42 °C when the OD 600 reached 0.6, and the procedure was monitored by the optical densities. The lysis rate based on clony-forming unit (CFU) and morphous of rSOBGs were detected as described previously 27 . Subsequently, the rSOBGs were treated with distilled water with streptomycin to kill the survival bacteria. O157:H7 EDL933 strain without pOSAmB plasmid were prepared to BGs as control (named OBGs).

Analysis of antigens by FACS. PBS contains chicken anti-Stx1B immunoglobin Y (IgY), rabbit
anti-Stx2A sera 32 or rabbit anti-intimin sera 33 was incubated with rSOBGs for 1 h, respectively. Then, the FITC-labeled rabbit anti-chicken IgG or goat anti-rabbit IgG was added and incubated for another 1 h as second antibody according to the first antibody, respectively. After processing, the labeled rSOBGs were incubated with 10 6 Hep-2 cells for 4 h. Finally, these cells were harvested and re-suspended with FACS buffer, and 3 × 10 4 Hep-2 cells were detected by flow cytometry. OBGs cells were used as control.
Mice immunization and challenge. Three-week-old male BALB/c mice (Vivarium of Academy of Military Medical Sciences, Beijing, China) were divided into groups randomly (Table S2). The mice were intragastricly administrated with 0.1 mg (corresponding to 10 8 CFU) rSOBGs, OBGs or PBS which were re-suspended in 100 μ l PBS on day 0 for primary and on day 14 for boost vaccinations, respectively. Different schemes for challenge were performed 14 days after the last immunization. 1. 25 mice from each group were intragastric challenge of 100 times 50% lethal doses (LD50, 10 9 CFU) viable E. coli O157:H7 EDL933; 2. 25 mice from each group were intragastric challenge of 500 LD50 viable E. coli O157:H7 EDL933; 3. 15 mice from each group were intraperitoneal injection with 2 LD50 (5 × 10 8 CFU) lysed E. coli O157:H7 88321; 4. 15 mice from each group were intraperitoneal injection with 5 LD50 lysed E. coli O157:H7 88321. The mice in scheme 1 and 2 (challenged with viable pathogenic bacteria) were deprived of food and water for 24 h and injected i.p. with Mitomycin (2.5 mg/kg).
In addition, 15 mice from each group were used to collect serum and irrigating solution of intestinal tract on days 0, 7, 14, 21 and 28 (3 mice were adopted for each time point) post the last immunization to determine the levels of specific antibodies. Surplus 10 mice from each group were used to collect the sera 14 days post the last immunization to perform the neutralization test in vitro. All remained mice were killed on day 21 post challenge. . Verification of the surface antigens. Flow cytometry was adopted to detect 3 × 10 4 Hep-2 cells. BGs were used as antigens, PBS contains chicken anti-Stx1B IgY, rabbit anti-Stx2A sera or rabbit anti-intimin sera was used as first anitibody, respectively. HRP labeled anti-chicken IgG or anti-rabbit IgG was used as test antibody according to their sources. (B). Cytotoxicity analysis of rSOBGs. Serial diluted BGs were incubated with Vero cell monolayer for 36 h, and the MTT assay was performed to detect the cytotoxicity. *P < 0.01 vs PBS group.
Examination of histopathology. The dead and survival mice in rOBGs group which were used for challenge were sacrificed and subjected to a full necropsy. Tissue specimens (liver, kidney and intestine) were collected for histologic examination. The specimens were fixed in 10% buffered neutral formalin and processed via standard procedures. Sections (5 μ m thickness) of paraffin-embedded tissues were stained with hematoxylin and eosin (H&E), and examined under optical microscopy.

Statistical analysis.
The results are presented as means ± standard error (S.E). Statistical significance was determined by ANOVA or Kruskal-Wallis test for parametric or nonparametric data, respectively. Mice survival differences were done by one sided Fisher's exact test. P values < 0.05 were considered significantly.

Results
rSOBGs expressed the linear Stx2Am-Stx1B fusion antigen on the surface of outer membrane. The expression plasmid pOSAmB containing 2300 bp OSAmB gene was constructed, and the co-transformation of plasmids pOSAmB and pLysE generated the vaccine strain of rSOBGs successfully.
The OD 600 reduced constantly after the shift of temperature. The precipitation 1 h after induction was harvested to evaluate the lysis rate, and it was counted as 99.99% ± 0.01%.
Outer membrane protein intimin could be confirmed in both rSOBGs and OBGs by flow cytometry, but the Stx2A and Stx1B were detected only on the surface of rSOBGs (Fig. 1A). rSOBGs were safe on cells model. At 10 7 CFU, the pathogenic bacteria killed nearly 100% Vero cells. In contrast, no obvious cytotoxic effect were detected when the Vero cells were treated with 10 mg (equivalent to 10 10 CFU of bacteria) rSOBGs or OBGs (P > 0.05) (Fig. 1B).
Immunization with rSOBGs induced various specific antibodies efficiently. Samples collected on days 0, 7, 14, 21 and 28 were detected. As same as our previous study (27), the specific antibodies induced by OBGs and rSOBGs increased persistently. In this manuscript, only the data on day 28 were displayed.
The antibodies specific to BGs were tested at first. IgA and IgG antibodies specific to BGs were elicited both by rSOBGs and OBGs (P < 0.01), and the levels of titer were similar (P > 0.05) ( Fig. 2A). No specific antibodies were detected in PBS group.
The antibodies specific to intimin, one of the most important factor in adhesion of EHEC, were tested subsequently. In contrast to PBS, both of the rSOBGs and OBGs could induce significant intimin-specific IgA/IgG antibodies (P < 0.01), and only one difference of IgG titers between rSOBGs and OBGs was detected in irrigation solution (P < 0.01) (Fig. 2B). The effect to induce Stx1-specific or Stx2-specific IgA/IgG antibodies will be the unique advantage of rSOBGs against OBGs in our expectation. In contrast to the OBGs and PBS, immunization of rSOBGs stimulated significant specific IgA antibodies against Stx1 and Stx2 in sera and irrigating solution (P < 0.01), and the specific IgG antibodies against Stx1 and Stx2 were detected only in sera (P < 0.01) (Fig. 2C,D). rSOBGs provided stronger and cross protection. rSOBGs not only displayed stronger protection against the viable E. coli O157:H7, but also displayed cross protection against the lysed E. coli O157:H7. rSOBGs provided the similar (P > 0.05) protection rate (22/25, 88%) with the OBGs (20/25, 80%) when the challenge dose was 100 LD50 (Fig. 3A). When the intragastric challenge dose was lifted to 500 LD50, the protection rate of OBGs (3/25, 12%) decreased sharply, and the rSOBGs provided higher protection rate (13/25, 52%) than that of OBGs (P < 0.01) (Fig. 3B). No mice survived in PBS groups when challenged either the high or the low dose. No mice survived in PBS groups after the challenge. The rSOBGs provided higher protection rate (11/15, 73.33%) than that of OBGs (P < 0.01) when challenge with 2 LD50 lysed E. coli O157:H7 88321 (Fig. 3C). When the dose upgraded to 5 LD50, 3 mice could survive in rSOBGs group but no survival in OBGs group (Fig. 3D).

Administration of rSOBGs reduced the damage of organs. The intragastric challenge with viable
E. coli O157:H7 EDL933 caused the major injury in intestine (epithelial cells necrosis), mainly accompanied with the damage of kidney, and no obvious damage was detected in liver (Fig. 4B). The immunization of rSOBGs reduced the damage in intestine efficiently, and the symptom in kidney vanished (Fig. 4C). The intraperitoneal injection with lysed E. coli O157:H7 88321 damaged the kidney (glomcrulus hyperaemia and death) and liver (focal necrosis and mesenchyme hyperaemia), but not the intestine (Fig. 4D). The administration of rSOBGs relieved these symptoms enormously (Fig. 4E). The normal mice were used as control (Fig. 4A).
The pathological changes of dead mice between OBGs group and rOBGs group were identical, and all of the survival mice showed negative serious pathological damage (data not shown). rSOBGs immune sera possessed neutralizing ability against lysed E. coli O157:H7. To test the neutralizing ablity of rSOBGs immune sera, it was mixed with lysed pathogen and injected i.p. mice . Ten percent mice (1/10) survived in the OBGs and PBS groups with 2 LD50 lysed E. coli O157:H7 88321. The rSOBGs provided significantly higher protection (80%, 8/10) than that of OBGs (10%, 1/10) (Fig. 5A). When the dose of pathogenic bacteria added to 5 LD50, all mice died in OBGs and PBS groups, but 2 mice(20%, 2/10) in rSOBGs group still survived (Fig. 5B).

Discussion
The 325 residue OmpA protein has been shown by physical analysis that N-terminal half of the molecule forms a β -barrel structure within the hydrophobic membrane layer that supports four loops (L1-L4) of amino acids at the outer surface, which may serve as OmpA receptor sites for the attachment of various phages and colicins [34][35][36] . In a previous study, a sandwich vector pHS64 which contained a recombinant (S. dysenteria and E. coli) OmpA had been constructed 37 . The OmpA sequence of S. dysenteria encodes five additional codons in L3 than E. coli, and these additinal codons are beneficial to the folding of foreign proteins. Some proteins, such as hemagglutinin 38 , VP1 39 , malarial's SERP antigen 40 , and the HBcAg-149 41 had been displayed in L3 by this vector.
rOSBGs could induce Stxs-specific antibodies in sera due to the surface expression of Stxs antigens, but failed to induce Stxs-specific IgG antibodies in irrigating solution, the interruption of intestinal mucosa against IgG antibody 42 and the low titer of IgG antibody may be the reseaons. The Stxs-specific antibodies could neutralize the Stxs, and the antisera of rOSBGs could provide effective cross protection against the challenge of low lethal dose lysed bacteria which contained two subtypes of Stxs. While this protection was locked into a narrow range of lethal dose challenges compared to the toxoid SAmB vaccine 13 . The decreased protection rate when challenged with high lethal dose may due to the low titer of the Stxs-specific antibodies which were induced by the insufficient expression of toxoid on the surface of E. coli O157:H7. The co-delivery of toxoid which were encapsulated by the BGs will be a beneficial attempt.
The intimin is the integrated membrane component on the surface of the E. coli O157:H7, and the levels of intimin specific antibodies hase been demonstrated to be the same between OBG and rSOBG due to the identical membrane structure. In addition, a higher intimin-specific IgG antibody was induced by rOSBGs than OBGs in sera irrigation solution. The adjuvant property of Stx1B may be a potential reason 43 . The promotion of adhesion between O157:H7 and target cells by Stx2 44,45 may be anorther reason. We presumed that the expression of SAmB may promote the adhesion between rSOBGs and enterocytes, more intimin could be displayed as antigens, and the higher titers of antibodies could be induced.
For the protection rate against viable E. coli O157:H7, although no significant difference was found between rOSBGs group and OBGs group when challenged with low dose (100 LD50) viable E. coli O157:H7, but the rOSBGs provided stronger protection than OBGs when challenged with 500 LD50 viable E. coli O157:H7. The adjuvant property of Stx1B, the adhesion promoting effect of Stx2 and the antitoxin function of rOSBGs may co-contributed to the protection.
In our previous study, the immunogenicity and protection effect of recombinant Stx2Am-Stx1B has been detected, could induce cross-neutralizing antibodies and effective protection 13 . Many similar constructs using the A or B subunit of Shiga toxin (Stx) elicited both bactericidal and toxin-neutralizing antibodies in mice. So far there is no clinical study of Stx-based human vaccine. In this study, we aimed to develop a new high effective oral vaccine against EHEC infection, which may easy accept to human. And Bacterial ghost (BG) based candidate vaccine have been used in various pathogen [23][24][25][26][27] , which show high sevurity and effectiveness. Thus, rOSBGs may be a better choice to children against EHEC.
Attaching/effacing and Shiga toxins are two important factors in infection of E. coli O157:H7. The co-infection of viable organisms and toxins is the general case in clinic examination 46 . In this study, the histopathology also revealed that the challenge of lysed O157:H7 caused damage of kidney (Stxs lesions), and the challenge of viable O157:H7 caused attaching and effacing lesions. E. coli O157:H7 infection characteristically results in attaching and effacing (A/E) lesions and toxin-induced systemic injury. Therefore, the novel vaccine candidate rSOBGs, which could provide co-prevention against the adhesion of bacteria and damage of Stxs, is a promising vaccine in human.