Salmonella enterica serovar Choleraesuis vector delivering SaoA antigen confers protection against Streptococcus suis serotypes 2 and 7 in mice and pigs

Streptococcus suis is one of the major pathogens that cause economic losses in the swine industry worldwide. However, current bacterins only provide limited prophylactic protection in the field. An ideal vaccine against S. suis should protect pigs against the clinical diseases caused by multiple serotypes, or at least protect against the dominant serotype in a given geographic region. A new recombinant Salmonella enterica serotype Choleraesuis vaccine vector, rSC0011, that is based on the regulated delayed attenuation system and regulated delayed antigen synthesis system, was developed recently. In this study, an improved recombinant attenuated Salmonella Choleraesuis vector, rSC0016, was developed by incorporating a sopB mutation to ensure adequate safety and maximal immunogenicity. In the spleens of mice, rSC0016 colonized less than rSC0011. rSC0016 and rSC0011 colonized similarly in Peyer’s patches of mice. The recombinant vaccine rSC0016(pS-SaoA) induced stronger cellular, humoral, and mucosal immune responses in mice and swine against SaoA, a conserved surface protein that is present in many S. suis serotypes, than did rSC0011(pS-SaoA) without sopB or rSC0018(pS-SaoA), which is an avirulent, chemically attenuated vaccine strain. rSC0016(pS-SaoA) provided 100% protection against S. suis serotype 2 in mice and pigs, and full cross-protection against SS7 in pigs. This new vaccine vector provides a foundation for the development of a universal vaccine against multiple serotypes of S. suis in pigs.


Introduction
Streptococcus suis (S. suis) strains are classified into 35 serotypes based on the capsular polysaccharide antigens. Globally, the predominant S. suis serotype isolated from clinical cases of the disease in pigs is serotype 2, followed by serotypes 9, 3, 1/2, and 7, together with 15.5% nontypable strains [1]. To date, only vaccination with S. suis bacterins is performed to prevent S. suis infection in swine. Although several studies have demonstrated homologous protection [2], the procedure usually has a high rate of failure because the antigenicity of the vaccine is damaged by heat and formalin processing, which leads to the production of antibodies that are not associated with protection and/or are cross-reactive [3,4]. Current vaccine research for S. suis focuses on subunit vaccine based on conserved proteins among different serotypes to protect pigs against clinical diseases caused by various S. suis serotypes, or protect against the dominant serotype in a given geographic region. However, the cost of many vaccines made from conserved proteins in veterinary medicine and swine medicine is relatively high compared to traditional bacterins.
To date, at least 24 subunit vaccine candidates have been identified [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22]. Most of them were tested in mouse models [5,8,11,12,15,[17][18][19][20][21][22], and only a few were evaluated in pigs [5,7,9,10,12,14,16,19,23,24]. The results showed that the protection conferred in swine is usually lower than that conferred in mice [10,23]. Therefore, it is necessary to evaluate the candidate Open Access *Correspondence: hyshi@yzu.edu.cn 1 College of Veterinary Medicine, Yangzhou University, Yangzhou 225009, Jiangsu, China Full list of author information is available at the end of the article antigen in the target animal, pig. Moreover, the capacities to induce cross-protection were evaluated for a few antigens. Of these antigens, the surface-anchored protein (Sao) is a membrane-anchored bacterial protein reacting with 28 of 33 S. suis serotypes and 25 of 26 serotype 2 isolates, suggesting that it is a highly conserved protein in S. suis [16]. The effectiveness of Sao as a vaccine candidate depends upon the adjuvant used. Sao formulated with Emulsigen-Plus ® provided no protection against S. suis serotype 2 (SS2) because the induced antibodies lacked opsonizing activity against SS2 [16], whereas Sao formulated with the Quil-A adjuvant partially protected pigs against aerosol challenge with SS2 because it induced opsonizing antibodies [24]. So far, Sao is the only protein that has been shown to induce cross-protection against serotype 1 and 7 strains in mice and pigs [19]. These results indicate that the efficacy of the protection induced by Sao as a subunit vaccine candidate may depend on the choice of appropriate vector or adjuvant.
Streptococcus suis is an encapsulated microorganism. Host protection against infection by S. suis is primarily mediated by opsonophagocytosis, which is mainly associated with a Th1-type immune response [25]. Salmonella has excellent adjuvant attributes and induces high mucosal, cellular and humoral responses [26]. However, live attenuated Salmonella, traditionally generated by deletion of its virulence genes, can lack a balance between the loss of its disease-causing ability and the ability to persist and induce immune responses [27], and is therefore not an ideal vaccine vector. In addition, constitutive high-level antigen synthesis causes a metabolic burden to the vaccine vector strain that can reduce the vaccine strain's ability to interact with host lymphoid tissues, resulting in a compromised immune response [28].
To solve these problems, the regulated delayed attenuation system was developed to make a strain displaying features of wild-type virulent strains of Salmonella at the time of immunization to enable strains to effectively colonize lymphoid tissues and then exhibit a regulated delayed attenuation in vivo to preclude inducing disease symptoms [29]. The regulated delayed antigen synthesis system could regulate antigen gene expression and permit high levels of antigen synthesis only after the vaccine strain reaches its target tissues [30]. The regulated delayed attenuation strategy includes a smooth-to-rough phenotypic change in lipopolysaccharides (LPS) with the presence or absence of mannose [31], and the replacement of the promoters of some virulence genes (fur, crp, and phoPQ) with a tightly regulated araC P BAD cassette so that the expression of these genes is dependent on arabinose supplement during in vitro growth [29]. Following the colonization of the lymphoid tissues, these virulence-associated proteins cease to be synthesized because no mannose or arabinose is present in vivo [32]. Therefore, attenuation manifests gradually in vivo, precluding the induction of disease symptoms and inducing the desired antigen-specific immune responses [33]. Furthermore, an araC P BAD cassette was used to regulate the expression of chromosomal lacI repressor gene, which binds to P trc on an expression plasmid and blocks antigen synthesis in the presence of arabinose in vitro. Once Salmonella reaches arabinose-free, host immunocompetent sites, the concentration of LacI decreases with each cell division, allowing upregulation of antigen synthesis and induction of desired antigen-specific immune responses [30].
In a previous study, a recombinant attenuated Salmonella enterica serotype Choleraesuis (Salmonella Choleraesuis) vaccine strain rSC0011 with the regulated delayed strategy in the wild-type Salmonella Choleraesuis C78-3 background was constructed [34]. Strain rSC0011, carrying 6-phosphogluconate dehydrogenase (6-PGD) gene of SS2, conferred 100% protection against an SS2 challenge in mice [34], but less than 50% protection in pigs (unpublished data). To increase the efficacy of our vaccine strain, a sopB mutation was introduced into rSC0011 vector to generate a new recombinant attenuated Salmonella Choleraesuis strain rSC0016 based on the previous report that the sopB mutation could improve the immunogenicity of Salmonella vaccine vector in mice [35]. The heterologous antigen gene, saoA from S. suis, was cloned into the Asd + expression plasmid pYA3493, generating plasmid pS-SaoA, which was used to transform the vector strain rSC0016, generating the candidate vaccine rSC0016 (pS-SaoA). The virulence and immune attributes of rSC0016 (pS-SaoA) in mice and piglets, and its crossprotection against SS2 and SS7 in piglets were evaluated. Our results showed that the improved recombinant attenuated Salmonella Choleraesuis, which combines the regulated delayed strategy and the sopB deletion, is more immunogenic than the parent strain rSC0011. It induces cross-protection against multiple serotypes of S. suis when delivering the heterologous protective antigen Sao. This could be a candidate for a universal vaccine against S. suis.

Ethics statement
All animal experiments were approved by the Jiangsu Administrative Committee for Laboratory Animals (permission number SYXK-SU-2007-0005) and complied with the Jiangsu Laboratory Animal Welfare and Ethics guidelines of the Jiangsu Administrative Committee of Laboratory Animals.
To detect the effect of sopB deletion, the fluid secretion in rabbit ileal loops after inoculation with wildtype strain C78-3, rSC0017 (Salmonella Choleraesuis C78-3 with sopB mutation), vaccine strains rSC0018(pS-SaoA), rSC0011(pS-SaoA), and rSC0016(pS-SaoA) were compared as previously described [41]. Briefly, 38-weekold New Zealand White rabbits were fasted overnight and anesthetized with isoflurane by an endotracheal tube. The ileum was exposed and ligated into several loops of 5-6 cm long with 1 cm spacers. 1 mL of indicated strains, containing around 1 × 10 9 colony-forming units (CFU), were injected into separate loops. LB medium was injected into one of the loops as a control. The abdominal musculature was closed using 3-0 chromic gut sutures and the skin closed with 3-0 ethilon sutures. Rabbits were maintained in a thermal blanket at 37 °C. After 8 h, the rabbits were euthanized with an overdose of sodium pentobarbital. The abdomen was reopened and the fluid within the ligated loops was collected and measured. The loops were fixed in 10% formalin and subjected to histopathological examination. All mutations were verified by PCR using corresponding primers. The ΔP crp527 ::TT araC P BAD crp deletion-insertion mutation was confirmed by growing on MacConkey maltose agar with or without 0.2% arabinose [29]. LPS profiles were examined by silver staining in 12% sodium dodecyl sulfate-polyacrylamide gel for Δpmi mutation [34]. LacI production for ΔrelA::araC P BAD lacI TT mutation was confirmed by Western blot [34]. The ΔasdA mutation was confirmed by growth with or without DAP in LB media [34].
Germany). The identity of the insert in pET28a was verified by DNA sequence analysis. Following digestion of plasmids pET28a-SaoA and pYA3493 with BamHI and SalI, the saoA from SS2 was cloned into pYA3493 to generate pS-SaoA. pS-SaoA and empty vector pYA3493 were transformed into the live attenuated Salmonella Choleraesuis ( Figure 1B).
To evaluate the production of SaoA in live attenuated Salmonella Choleraesuis, sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was performed as described [33]. Strains harboring plasmid pS-SaoA or control vector pYA3493 were grown in LB medium with indicating supplements (0.2% arabinose and mannose) at 37 °C with aeration. When the culture reached an OD 600 of 0.8, Isopropyl-d-thiogalactopyranoside (IPTG) was added to the culture, and the culture continued to grow for 3 h. The bacteria numbers were normalized by OD 600 . 1 mL of culture was collected for Western blot analysis using anti-SaoA antiserum as previously described [33]. Briefly, total protein samples were separated on a 10% (wt/vol) SDS-PAGE gel and transferred onto nitrocellulose membranes. Phosphate buffered saline (PBS) with 5% fat-free milk powder and 0.05% Tween 20 (PBS-T) was used for blocking. The membrane was incubated with an appropriate anti-rabbit polyclonal antibody (1:10 000, anti-SaoA) or anti-GroEL (Sigma, St. Louis, MO, USA) for 1 h at room temperature, washed three times with PBS-T, and then horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G was used as the secondary antibody (Sigma). The membrane was developed with a chemiluminescent substrate using the kit ECL Plus Western Blotting System (GE Healthcare, Chalfont St Giles, UK) according to manufacturer's instructions. Densitometry was quantified using Image J software (Image J2) [43].
To examine the stability of plasmids pS-SaoA and pYA3493 in Salmonella Choleraesuis vector, the strains containing pS-SaoA or pYA3493 were cultured with the daily passage of 1:1000 dilutions for 5 consecutive days (about 50 generations) in LB medium with DAP, arabinose, and mannose. At the 50th generation, Asd + plasmids pS-SaoA and pYA3493 were tested by endonuclease digestion. The mutations of the live attenuated Salmonella Choleraesuis vector were confirmed by PCR. SaoA production was evaluated by Western blot.

Preparation the SaoA protein of SS2 or SS7, antiserum to the SaoA protein of SS2, and Salmonella outer membrane proteins (SOMPs) of Salmonella Choleraesuis
Escherichia coli strain BL21 carrying pET28a-SaoA was used for the synthesis of the His-tagged SaoA fusion protein (Table 1). Cells were grown to mid-log phase (OD 600 of 0.6) in LB medium with kanamycin at 37 °C and induced with 1 mM IPTG for 3 h. The His-tagged SaoA protein was purified by using a CelLytic B Plus kit (Sigma) and a His-Select nickel affinity gel (Sigma) according to manufacturer's instructions. To create rabbit antibodies to SaoA, two female New Zealand White rabbits (8 weeks old) were injected subcutaneously with an emulsion consisting of a 1:1 ratio of 400 μg of the His-tagged SaoA protein to complete Freund adjuvant (Sigma). Two weeks after the primary injection, the immunization was repeated but with incomplete Freund adjuvant (Sigma) instead of complete adjuvant. Three weeks after the second immunization, the rabbits were boosted with 200 μg of the His-tagged SaoA protein without adjuvant. Two weeks after the third immunization, the rabbits were bled to obtain anti-SaoA rabbit antiserum [34].
Salmonella outer membrane proteins (SOMPs) were prepared from wild-type Salmonella Choleraesuis C78-3. Briefly, total envelope pellets were suspended in 4 mL of 20 mM Tris-HCl (pH 8.6) containing 1% Sarkosyl and incubated for 30 min on ice. The outer membrane fraction was obtained as a pellet after centrifugation at 132 000 × g at 4 °C for 1 h. The pellet was re-suspended in 4 mL of 20 mM Tris-HCl buffer (pH 8.6) and stored at −20 °C.

Distribution of Salmonella bacteria in BALB/c mice
Colonization assay for mutant strain was carried out as described [34]. 140 7-week-old female BALB/c mice were divided into seven groups with 20 mice per group. Groups of mice were orally inoculated with 20 μL buffered saline with gelatin (BSG) containing 1 × 10 9 CFU of Salmonella strains. Liver, spleen, and Peyer's patches of mice were collected on days 3, 7, 14, and 21 post-infection. Tissues were weighed and homogenized in a final volume of 1 mL BSG [39]. Serial dilutions were plated onto MacConkey agar plates containing 1% lactose, with or without 0.2% arabinose and 0.2% mannose, to determine the number of viable bacteria. Plates were incubated at 37 °C for at least 18 h. The residual 900 mL of homogenized tissues were inoculated into 5 mL Tetrathionate Broth (Difco) for Salmonella enrichment when no colonies were observed on the plates. Samples that were negative by direct plating and positive by enrichment were recorded as 10 CFU/g. Samples that were negative by both direct plating and enrichment were recorded as 0 CFU/g [39]. The assay was performed twice, and the data were pooled.

Immunization of mice
Salmonella Choleraesuis vector strains harboring plasmids pS-SaoA or pYA3493 (empty vector) were grown in LB broth with 0.2% arabinose and 0.2% mannose overnight at 37 °C as standing cultures that were diluted 1:100 in the same medium. The culture was grown with aeration (180 rpm) at 37 °C to an OD 600 of 0.85-0.9. Bacteria were collected by centrifugation at room temperature and re-suspended in BSG. 15 7-week-old BALB/c female mice were inoculated orally with 20 μL BSG containing 1 ± 0.3 × 10 9 CFU of Salmonella Choleraesuis vectors with pS-SaoA or plasmid pYA3493 or 20 μL BSG control. Mice were boosted with the same dose of the same strain 3 weeks later. About 100 μL of whole blood was obtained by mandibular vein puncture 3 weeks after primary inoculation and 2 weeks after boosting. Serum was removed from the whole-blood samples and stored at −20 °C. Vaginal-wash samples in mice were collected at the indicated time and stored at −20 °C as described [33,34]. Ten mice from each group were challenged with SS2, and five mice were used for cytokine detection. This experiment was performed twice, with each group (15 mice) receiving approximately the same dose of vaccine. The results from both experiments were similar and the data were pooled for analysis.

Immunization of piglets
The preparation of the inoculum of Salmonella Choleraesuis vector strains harboring plasmids pS-SaoA or pYA3493 was performed as described in a previous section. Seventy-five 3-week-old castrated piglets were purchased from a commercial pig farm in Jiangsu Province, China. To select the pigs with negative antibody titers against SS2, SS7, and Salmonella Choleraesuis, blood was obtained from pigs by cranial vena cava and incubated at 37 °C for 60 min. The resulting clot was pelleted by centrifugation. Serum was removed from the whole-blood sample and stored at −20 °C. Serum IgG responses to SS2 SaoA, SS7 SaoA, and Salmonella Choleraesuis outer membrane proteins (OMPs) were measured by enzymelinked immunosorbent assay (ELISA) (see below) [33,44,45]. The IgG titer that was less than cut-off value was considered as seronegative. These piglets were randomly assigned to five groups with 15 piglets/group, named rSC0016(pS-SaoA) group, rSC0016(pYA3493) group, rSC0018(pS-SaoA) group, rSC0018(pYA3493) group, and BSG group. Piglets were inoculated orally with 10 mL 1 ± 0.3 × 10 9 CFU of Salmonella Choleraesuis vector strains carrying either the SaoA expression plasmid pS-SaoA or empty plasmid pYA3493 or BSG control. Piglets were boosted with the same dose of the same strain 3 weeks later. Serum and nasal swabs were collected at 3, and 5 weeks after first immunization [10,16]. In each group, five piglets were challenged with SS2 and five with SS7 at 2 weeks after a boost. Five piglets were used to detect cytokine levels in the blood and spleen at different time points (0.5, 3, 5, and 7 days after boost). This experiment was performed twice with each group receiving approximately the same dose of vaccine. The results from both experiments were similar and the data were pooled for analysis.

ELISA test
Serum IgG responses to SaoA, to OMPs of Salmonella Choleraesuis in mice and pigs, and vaginal wash IgA antibody to SaoA in mice and nasal cavity wash IgA in pigs against SaoA were measured by ELISA. The initial dilutions of serum or vaginal/nasal wash from an individual animal were 1:50 or 1:10, respectively. Briefly, 100 µL solutions containing either 1 μg/well of Salmonella OMPs or of SaoA in sodium carbonate-bicarbonate coating buffer (pH 9.6) was used to coat Nunc-Immuno MaxiSop 96-well plates (Corning, NY, USA). The plates were then incubated overnight at 4 °C. The next day, the plates were washed 3 times with PBST (PBS with 0.1% Tween 20) and then blocked with a PBST containing 2% BSA solution for 2 h at room temperature. A 100 μL volume of serially diluted sample was added in triplicate to individual wells, and the plates were incubated for 1 h at room temperature. After washing wells with PBST, goat anti-mouse or goat anti-pig IgG-horseradish peroxidase (HRP) detection antibody (Sigma, St. Louis, MO, USA) diluted 1:5000 in PBST was added to the wells and the plate was incubated at room temperature for 60 min with gentle rocking. Plates were developed with 1-Step TM Ultra TMB-ELISA (Thermo Fisher Scientific, Waltham, MA, USA) and quenched with 3N H 2 SO4. Absorbance was recorded at 450 nm using an automated ELISA plate reader (model EL311SX; Biotek, Winooski, VT). Absorbance readings 2.1 times higher than the baseline values of preimmune sera (negative control: about 0.045-0.052) were considered positive [33,[44][45][46]. Sera were collected from mice or piglets on 0.5, 3, and 5 days. In addition, spleen tissues were removed at 7 days after boost immunization. Spleens were homogenized in a final volume of 1 mL (mouse tissues) or 5 mL (pig tissues) BSG at a ratio of weight: volume of 1:9. The supernatants were stored at −70 °C. Cytokines in sera and spleen of mice or pigs were analyzed by sandwich ELISA using commercial kits (Abcam, Cambridge, UK, except pig-specific IL4 ELISA kit, which was purchased from Thermo Fisher Scientific) according to manufacturer's protocols.

Analysis of the cross-reactivity between SS2 and SS7 in pigs
Since the saoA gene was from SS2, the cross-reactivity of sera from pigs immunized with rSC0018(pS-SaoA) or rSC0016(pS-SaoA) was checked against total cell lysates from SS2 or SS7, respectively. Total S. suis proteins (SS2 or SS7 strain) were extracted from bacterial culture and 10 μg of proteins were separated by 12% SDS-PAGE gel electrophoresis and transferred to nitrocellulose membrane. The membrane was incubated with immunized pig serum (1:100 dilution) for 1 h at room temperature and subsequently washed three times with 0.05% w/v Tween-20 in TBS for 5 min. Anti-pig IgG conjugated to horseradish peroxidase (1:5000 dilution, Sigma) was used as secondary antibodies. The membrane was developed with a chemiluminescent substrate using the kit ECL Plus Western Blotting System (GE Healthcare) according to manufacturer's instructions.

Challenge experiment with SS2 in mice or SS2 and SS7 in pigs
Ten mice were challenged with 3 × 10 8 CFU of SS2 intraperitoneally (i.p.) at 2 weeks after the booster. The 50% lethal dose (LD 50 ) of SS2 in BALB/c mice was 1.2 × 10 7 CFU. Challenged mice were monitored daily for 15 days [33].
Five pigs in each group were challenged with SS2 by intravenous injection (i.v.) with 5.0 × 10 8 CFU (LD 50 of SS2 in pigs was 5.6 × 10 7 CFU). The clinical symptoms and survival of pigs were monitored for 14 days. When needed, the brain of pigs was fixed in 10% formalin buffer (pH 7.2). Thin sections (5 mm) of tissue were stained with hematoxylin/eosin and examined by light microscopy to detect histological changes caused by SS2. Another five pigs in each group were challenged by i.v. with 1.0 × 10 10 CFU of SS7. The rectal temperatures of pigs challenged with SS7 were measured at 0, 1, 2, 3, 4, 5, 6, and 7 days post-inoculation [47].

Statistical analysis
Data were presented as the geometric means and standard deviations for all assays. The Mann-Whitney U test (Graph-Pad Software, Inc., San Diego, CA) was used to evaluate the persistence of strains in vivo and antibody levels of mice. The Kaplan-Meier method (SPSS software) was applied to obtain the survival fractions following i.p. challenges of immunized mice and i.v. challenges of immunized pigs. A P value of 0.05 was considered statistically significant.

Construction of sopB deletion in Salmonella Choleraesuis vaccine vectors
Previous reports have shown that the inactivation of sopB in Salmonella Typhimurium improves humoral and cellular immunity in the host [35,48]. To improve the safety and increase the immunogenicity of Salmonella Choleraesuis vector, a sopB mutation was introduced into wildtype Salmonella Choleraesuis C78-3 to generate strain rSC0017 ( Figure 1A) or into Salmonella Choleraesuis rSC0013 to generate strain rSC0016 (Table 1). Strain rSC0016 has the mutations for the regulated delayed attenuation system and the regulated delayed antigen synthesis system, characterized by ΔP crp ::TT araC P BAD crp, Δpmi, and ΔrelA::araC P BAD lacI TT [34], in addition to the sopB mutation. It also has a ΔasdA mutation for the balanced-lethal system to facilitate its use as a vector.

Phenotypic characterization of live attenuated Salmonella Choleraesuis vaccine vector with sopB mutation
SopB plays a role in water efflux, fluid secretion, and subsequent diarrhea associated with Salmonella infection [49]. In this study, the vaccine strains rSC0018(pS-SaoA), rSC0011(pS-SaoA), rSC0016(pS-SaoA), the strain rSC0017 with a sopB deletion mutation, the parental wild-type strain C78-3, and a LB control were injected into six parts of the ileal loop in rabbits to detect the effect of the sopB deletion in Salmonella Choleraesuis. Hematoxylin-eosin staining of ileal loop sections revealed populations of neutrophilic cells in the ileum loops infected by the wild-type strain C78-3, but not in those infected by rSC0017, rSC0018(pS-SaoA), rSC0011(pS-SaoA), or rSC0016(pS-SaoA). These loops were similar to the loops injected by LB ( Figure 1C). The fluid accumulation in the ileal loops injected with the wild-type parental strain C78-3 was significantly higher than those injected with rSC0017, rSC0018(pS-SaoA), rSC0011(pS-SaoA), rSC0016(pS-SaoA) and LB (Figure 1D). Particularly, the inflammatory exudate in the ileal loops injected with rSC0016(pS-SaoA) containing sopB mutation was significantly reduced compared to that injected with rSC0011(pS-SaoA) without the sopB mutation. In addition, the fluid accumulation in the ileal loops injected with rSC0018(pS-SaoA) was significantly more than that in those injected with rSC0011(pS-SaoA) and rSC0016(pS-SaoA), even higher than that of rSC0017 ( Figure 1D). Little fluid was detected in the loops injected with LB control, which was significantly less than those injected with C78-3, rSC0017, SC0018(pS-SaoA), rSC0011(pS-SaoA), and rSC0016(pS-SaoA) ( Figure 1D).
The phenotype of ΔP crp527 ::TT araC P BAD crp mutation was tested on MacConkey maltose agar with and without arabinose. The strain with ΔP crp527 ::TT araC P BAD crp mutation formed red colonies on MacConkey maltose agar in the presence of arabinose, and white colonies in the absence of arabinose (Figure 2A) [29,46]. The smooth LPS pattern was observed in strain rSC0016 grown with mannose, and rough without mannose due to the presence of Δpmi mutation ( Figure 2B) [34]. To verify the regulated delayed synthesis of heterologous protein, strain rSC0016(pS-SaoA) was passaged four times in arabinose -and mannose-free medium after it was grown in nutrient broth medium with 0.2% arabinose and mannose, as described previously [34]. The expected phenotype, the gradual disappearance of the LacI protein and the increased production of the heterologous protein SaoA, was observed in the strain rSC0016(pS-SaoA) ( Figure 2C) [29,46]. The introduction of the ΔasdA mutation into Salmonella Choleraesuis strains rSC0010, rSC0013, and C500 (Table 1) [46,50] did not affect the growth of the resulting strains rSC0011, rSC0016, and rSC0018 ( Figure 2D). However, the production of SaoA protein in rSC0011 and rSC0016 was 2.2-fold and 4.7fold higher than that in rSC0018, respectively, whereas the production of SaoA protein in strain rSC0016 was 2.5-fold higher than that in strain rSC0011(pS-SaoA) ( Figure 2E).

Plasmids pYA3493 and pS-SaoA are stable in attenuated Salmonella Choleraesuis rSC0016 strain
A single colony of rSC0018, rSC0016 and rSC0011 either with pS-SaoA or pYA3493 was grown in LB with 0.2% mannose, 0.2% arabinose, and 50 µg/mL DAP for 50 generations [34]. All the colonies that were examined using PCR or endonuclease digestion contained either Figure 2 Phenotypic characterization of Salmonella Choleraesuis strains rSC0016. A Phenotype of ΔP crp527 ::TT araC P BAD crp. Strain rSC0016 with the ΔP crp527 ::TT araC P BAD crp mutation were grown on MacConkey maltose agar with and without 0.2% arabinose. B Phenotype of Δpmi mutation. LPS profiles of Δpmi mutant strains rSC0016 in NB grown with or without 0.2% mannose. Lanes: 1, wild-type C78-3; 2, rSC0016 with mannose; 3, rSC0016 without mannose. C Regulated decreased synthesis of LacI and regulated delayed synthesis of SaoA proteins in rSC0016(pS-SaoA) containing ΔrelA::araC P BAD lacI TT mutation. Strain rSC0016(pS-SaoA) were grown in NB with arabinose and mannose (Lane 1) and then diluted 1:10 into fresh NB without arabinose and mannose until OD 600 to 0.8. The process was repeated for four times (Lane 2-5); each lane was loaded around 2.5 × 10 7 CFU bacteria. Synthesis of LacI and SaoA were detected by Western blot using correspondent antiserum. M: protein marker. D Phenotype of ΔasdA mutation. Growth curves of rSC0016 in LB with and without DAP. Growth was monitored by measuring OD 600 at the indicated time intervals. E Synthesis of SaoA in Salmonella Choleraesuis vector rSC0011, rSC0016 and rSC0018. An equal amount of cells was subjected to SDS-PAGE analysis. Immunoblot was detected with SaoA-specific polyclonal antibody. Densitometry ratio was quantified using Image J software. M: protein Marker; 1: rSC0011(pS-SaoA); 2: rSC0016(pS-SaoA); 3: rSC0018(pS-SaoA); 4: rSC0011(pYA3493); 5: rSC0016(pYA3493); 6: rSC0018(pYA3493); GroEL was used as a control. plasmid pS-SaoA or plasmid pYA3493, indicating that pS-SaoA and pYA3493 are equally stable in the vaccine strains (data not shown). These colonies can produce the expected SaoA protein (data not shown).

Salmonella Choleraesuis vaccine vector with sopB mutation induces higher antibody responses in mice
To test the immunogenicity of Salmonella Choleraesuis vaccine strain that combines regulated delayed strategy and the sopB mutation, the antibody responses to SS2 SaoA and Salmonella Choleraesuis OMPs in mice were measured at 3 and 5 weeks after the primary immunization ( Figures 4A, B and C). Higher serum IgG and vaginal IgA titers against SaoA were detected in the mice immunized with the live attenuated Salmonella Choleraesuis vaccines carrying plasmid pS-SaoA [rSC0016(pS-SaoA), rSC0011(pS-SaoA), and rSC0018(pS-SaoA)] compared to those with BSG or the strain containing the empty vector pYA3493 (Figures 4A and B). The levels of serum IgG against SaoA in mice immunized with rSC0016(pS-SaoA) were similar to those in mice immunized with rSC0011(pS-SaoA), yet significantly higher than those immunized with rSC0018(pS-SaoA) ( Figure 4A). Higher vaginal IgA titers against SaoA were observed in the mice immunized with rSC0016(pS-SaoA) or rSC0011(pS-SaoA) compared to those with rSC0018(pS-SaoA) ( Figure 4B). At 3 or 5 weeks after vaccination, the mucosal IgA responses against SaoA in mice immunized with rSC0016(pS-SaoA) were also significantly higher than those with rSC0011(pS-SaoA) ( Figure 4B). At 3 weeks after the primary immunization, the IgG antibody responses to Salmonella Choleraesuis OMPs were significantly higher in mice immunized with rSC0016(pS-SaoA) than those with rSC0011(pS-SaoA) or rSC0018(pS-SaoA) ( Figure 4C). At 5 weeks postinfection, higher IgG responses against Salmonella Choleraesuis OMPs were observed in mice immunized with rSC0016 carrying either pS-SaoA or pYA3493 than those with rSC0011 or rSC0018 ( Figure 4C).

Salmonella Choleraesuis vaccine vector with sopB mutation induces higher interferon γ (IFN-γ), interleukin 4 (IL-4), and interleukin 17A (IL-17A) responses in mice
A previous study showed that the introduction of a sopB mutation into different attenuated Salmonella Typhimurium strains enhanced the immune responses to the delivered heterologous antigen in mice [48]. To further characterize the influence of sopB on immune responses, the levels of IFN-γ, IL-4, and IL-17A in mice sera collected at 0.5, 3, and 5 days and in the spleen collected at 7 days after the booster were measured. The concentrations of IFN-γ, IL-4, and IL-17A induced in mice immunized with vector rSC0016, rSC0011, or rSC0018 carrying either pYA3493 or pS-SaoA were significantly higher in the sera and spleen tissues than those induced in mice treated with BSG (Figures 5A-C, P < 0.01). The three vectors that contain pS-SaoA (rSC0011, rSC0016 and rSC0018) induced higher levels of IFN-γ, IL-4, and IL-17A than the strain with the empty vector pYA3493 in sera and spleen of mice (Figures 5A-C, *P < 0.05; **P < 0.01). Conversely, rSC0018(pS-SaoA) induced similar levels of IFN-γ and IL-4 in sera compared to those induced by rSC0018(pYA3493) at 3 and 5 days after boost immunization ( Figures 5A and B).
Higher IFN-γ, IL-4, and IL-17A levels were produced in the sera and spleens of mice immunized with rSC0016 or rSC0011 carrying either the empty vector pYA3493 or the SaoA-expressing plasmid pS-SaoA than in those with rSC0018 (Figures 5A-C; # P < 0.05, ## P < 0.01). rSC0011 with pYA3493 or pS-SaoA induced similar levels of IL-4 to those of rSC0018 with pYA3493 or pS-SaoA (Figure 5B). Higher levels of IFN-γ, IL-4, and IL-17A were observed in the sera and spleens of mice immunized with rSC0016(pS-SaoA) compared to those in mice immunized with rSC0011(pS-SaoA) (Figures 5A-C; # P < 0.05, ## P < 0.01). Alternatively, the level of IL-17A in the spleens of mice immunized with rSC0016(pS-SaoA) was similar to rSC0011(pS-SaoA) ( Figure 5C). The levels of IFN-γ, IL-4, and IL-17A that were induced by rSC0016(pYA3493) were significantly higher than those induced by rSC0011(pYA3493) in the sera at 3, and 5 days after the booster, as well as in the spleen (Figures 5A-C; # P < 0.05, ## P < 0.01). These results indicate that the vector rSC0016 with the ΔsopB mutation induced enhanced responses more than the rSC0011 strain did [35].

Salmonella Choleraesuis vaccine vector with sopB mutation induces higher levels of IFN-γ, IL-4, and IL-17A in piglets
To further characterize the immune trends in piglet immunized with strain rSC0016 delivering a heterologous antigen, the levels of cytokines IFN-γ, IL-4, and IL-17A in the sera of immunized piglets at 0.5, 3, and 5 days, and in the spleens at 7 days after a boost were tested ( Figures 7A-C). Significantly higher levels of IFN-γ, IL-4, and IL-17A were observed in the piglets immunized with the strains rSC0018 and rSC0016 carrying either pS-SaoA or the empty vector pYA3493 compared to those with BSG group (Figures 7A-C; P < 0.01). Higher levels of IFN-γ, IL-4, and IL-17A were also detected in the swine immunized with strains rSC0018 and rSC0016 carrying pS-SaoA compared to the strains carrying the empty vector pYA3493 (Figures 7A-C; *P < 0.05, **P < 0.01). Conversely, the levels of IFN-γ and IL-17A in the sera of swine immunized Choleraesuis OMPs were measured by ELISA. The data represent reciprocal antibody titers in sera from ten mice orally immunized with attenuated Salmonella carrying either pS-SaoA or pYA3493 (empty vector) and BSG at indicated weeks after immunization. Serum and vaginal wash obtained from individual mice were serially diluted to obtain titers, starting from either 1:50 or 1:10. Error bars represent variation between mice. Significant differences were indicated. **P < 0.01, for Salmonella Choleraesuis with pS-SaoA compared to Salmonella Choleraesuis with pYA3493 in A and C; # P < 0.05; ## P < 0.01, for Salmonella carrying either pS-SaoA or pYA3493 compared each other. No antibody responses were detected to antigen tested in mice immunized with the only BSG or in pre-immune sera from vaccinated mice. ELISA was performed twice with identical results. ▸ Figure 5 Cytokines levels in ten mice immunized with Salmonella Choleraesuis vaccines. IFN-γ (A), IL-4 (B) or IL-17A (C) in sera at 0.5, 3, and 5 days, in the spleen at 7 days after the boost were assayed with ELISA kit. BSG control was also included. *P < 0.05; **P < 0.01, for the cytokines levels induced by strains rSC0011, rSC0016 and rSC0018 containing pS-SaoA compared with that containing the empty vector pYA3493; # P < 0.05, ## P < 0.01 for the significant differences between groups were indicated. The assay was performed in triplicate. The data were collected from two experiments and analyzed.
with rSC0018(pS-SaoA) were equal to those immunized with rSC0018(pYA3493) at 3 and 5 days after the booster. Significantly higher levels of IFN-γ, IL-4, and IL-17A were detected in the piglets immunized with rSC0016 carrying either pS-SaoA or pYA3493 compared to those with rSC0018 (Figures 7A-C; # P < 0.05, ## P < 0.01), except in the serum at 12 h after the booster. Furthermore, the levels of IFN-γ and IL-17A in the swine immunized with rSC0016(pS-SaoA) were 50-fold higher in the sera 5 days after the booster compared to those with rSC0018(pS-SaoA) ( Figures 7A and D). IL-17, mainly produced by Th17 cells [53][54][55], can synergize with IFN-γ to enhance the production of proinflammatory cytokines [56]. Our results indicated that the new live attenuated strain rSC0016 is more immunogenic than the chemically attenuated strain rSC0018.

Salmonella Choleraesuis vaccine vector with sopB mutation confers protection against SS2 in piglets
Previous studies showed that the percentage of protection conferred by a specific vaccine in pigs is usually lower than that in mice [7,10,23]. Therefore, the protective efficacy of the live attenuated strain rSC0016(pS-SaoA) against SS2 was evaluated in piglets. It provided 100% protection against 25 times the LD 50 of SS2 challenge in mice. Following challenge with 5 × 10 8 CFU of SS2 (8.9 × LD 50 in piglets) via the ear vein, all the pigs immunized with either rSC0018(pYA3493) or BSG died within 48 h of challenge ( Figure 8A). The piglets immunized with rSC0016(pYA3493) survived longer but died at 84 h after challenge ( Figure 8A). The swine immunized with rSC0018(pS-SaoA) died 108 h after challenge and showed symptoms of breathing difficulty and joint swelling ( Figures 8A and B). In contrast, all the swine immunized with strain rSC0016(pS-SaoA) survived throughout a 14-day observation period and appeared healthy throughout this period, except for temporary depression within 48 h of a challenge. A histopathological analysis of the challenged swine immunized with either strain rSC0018(pS-SaoA) or BSG showed hemorrhage, congestion, and inflammatory exudation in the brain ( Figure 8B). The swine immunized with strain rSC0016(pS-SaoA) showed normal structures ( Figure 8B). These results suggest that strain rSC0016(pS-SaoA) provided excellent protection against SS2 in the piglet model.

Salmonella Choleraesuis vaccine vector with sopB mutation confers cross-protection against SS7 in piglets
Sao is a highly conserved protein in S. suis [16]. The protection against SS7, one of the main S. suis serotypes in China, conferred by rSC0016(pS-SaoA) was evaluated in piglets. SS7 is not lethal in pigs and only led to a fever in pigs after intravenous challenge. The cross-reactivity of the sera from pigs immunized with Salmonella Choleraesuis vector rSC0016, carrying SS2 SaoA protein, against SS7 SaoA protein was tested by Western blot. The results showed that the sera from the pigs vaccinated with rSC0016 or rSC0018, carrying SS2 SaoA protein, could recognize the SS7 SaoA protein as well as the SS2 SaoA protein ( Figure 8C).

Discussion
Recently, an innovative Salmonella Choleraesuis live attenuated vaccine vector rSC0011 with the mutations for regulated delayed attenuation (Δpmi-2426 deletion mutation, ΔP crp527 ::TT araC P BAD crp deletion-insertion mutation) [29,33], and regulated delayed antigen synthesis (ΔrelA::araC P BAD lacI TT deletion-insertion mutation) was developed [29,57]. This vaccine vector, delivering a conserved protein of S. suis, only conferred protection against a low-dose challenge of SS2 in mice [34], but not a high-dose challenge of SS2 (Table 2). Therefore, it was necessary to improve the immunogenicity of the strain, rSC0011. The SopB protein of Salmonella Typhimurium plays an immunosuppressive role [48], and its inactivation improves humoral and cellular immune responses of the host [35,48]. In this study, a sopB deletion mutation was introduced into our Salmonella Choleraesuis vector with the regulated delayed attenuation system to generate the new strain rSC0016. This new strain displayed the expected phenotypes of reduced fluid secretion and inflammation in rabbit ileal loop test, indicating that the ΔsopB deletion in Salmonella Choleraesuis had the same effect as it had in other Salmonella serotypes [52,58]. Strain rSC0016 showed a similar distribution to that of strain rSC0011 in Peyer's patches of mice, confirming that the sopB mutation did not affect the gut colonization of the strain [48,52]. However, the bacterial load of rSC0016 in the spleen was significantly lower than that of rSC0011. This result indicates that strain rSC0016 is more attenuated than rSC0011 in deep lymphoid tissue. Our results demonstrated that strain rSC0016, with the ΔsopB mutation, induces higher levels of IgA, IgG, IFN-γ, IL-4 γ and IL-17A in mice than the other strains tested. These results may relate to the immunosuppressive function of sopB. In addition, it might also be related to the higher antigen loads of rSC0016 in the tissues than those of rSC0018. When Salmonella enters the cell, 10-20% of the bacteria remain in the cytosol, whereas the majority of the internalized bacteria remain within a modified phagosome, the Salmonella-containing vacuole [59,60]. The cytosolic bacteria replicate faster than the vacuolar bacteria and are invasion-primed and -competent [60]. A small but significant amount of the cytosolic bacteria can leave epithelial cells, facilitating a rapid secondary infection and inducing both enteric and systemic infections [60]. The sopB deletion mutant does not affect the intracellular replication of the bacterium but does induce the early death of these cells, increases lysis, and promotes the transmission of the bacteria [59]. The premature death of epithelial cells exposes the strain to the immune system at a great efficiency, which may contribute to the increased immune responses triggered by rSC0016 compared to rSC0011. , serum IgG responses to Salmonella Choleraesuis OMPs (C) in ten pigs were measured by ELISA. The data represent reciprocal anti-IgG antibody titers from the piglets orally immunized with attenuated Salmonella carrying either pS-SaoA and pYA3493 at the indicated weeks after immunization. Serum and nasal cavity wash obtained from individual pig were serially diluted to obtain titers, starting from either 1:50 or 1:10. Error bars represent variation between different pigs. Significant differences were indicated. **P < 0.01, for Salmonella Choleraesuis with pS-SaoA compare to Salmonella Choleraesuis with pYA3493 in A and B; # P < 0.05; ## P < 0.01, for Salmonella carrying either pS-SaoA or pYA3493 compared each other. No responses were detected to antigen tested in pigs immunized with BSG or in pre-immune sera from vaccinated piglets. ELISA was performed twice with identical results.

Figure 7
Cytokine levels in ten pigs. IFN-γ (A), IL-4 (B) or IL-17A (C) in sera from 0.5, 3, and 5 days, in the spleen at 7 days after the booster from single piglet were assayed with ELISA. BSG controls were also included. *P < 0.05; **P < 0.01, for the cytokines levels induced by strains rSC0011, rSC0016 and rSC0018 containing pS-SaoA compared with that containing the empty vector pYA3493; # P < 0.05, ## P < 0.01 for significant differences between groups were indicated. The assay was performed in triplicate. The data were collected from two experiments and analyzed.
The Salmonella Choleraesuis C500 vaccine induces mucosal IgA and IFN-γ responses during its protection of pigs [61], whereas our strain rSC0016 induced greater IgA, IgG, IFN-γ, IL4, and IL17A responses compared to strain rSC0018, which was derived from C500. IL-17A induces and mediates proinflammatory responses and the expression of many other cytokine genes. The higher levels of IL-17A detected in the sera (34-50 fold) of swine immunized with rSC0016 compared to those with rSC0018 may help the swine clear the infecting pathogen and confer greater protective immunity. The detailed mechanism underlying high IL-17A levels induced by rSC0016 in pig warrants further study. The immune responses induced by strains rSC0016 and rSC0018 in both mice and weaned piglets were compared. In mice, the antibody titers of IgG and IgA against Sao of S. suis and IgG against the OMPs of Salmonella Choleraesuis induced by rSC0016 were significantly higher than those induced by rSC0011. Both were higher than those induced by rSC0018, indicating that the immune responses induced by the combination of the regulated delayed strategy and the sopB mutation were superior to those induced by the regulated delayed strategy alone or with chemically attenuated rSC0018. The trends in the immune responses of weaned piglets were similar to those of mice. It is noteworthy that the mucosal IgA responses induced by rSC0016 were significantly higher than those induced by rSC0011 in mice, and were significantly higher than those induced by rSC0018 in mice or in piglets after immunization. Our orally delivered Salmonella vaccine vectors induced strong mucosal and humoral responses against the Sao antigen (Figures 4 and 6). IgA could prevent the adhesion of pathogens to mucosal cells and improve the protection Figure 8 Protection in pigs. A Survival curve of swine after challenge with SS2. Groups of ten pigs were orally immunized twice at 3-week intervals with indicated strains and challenged intravenously with 10 8 CFU of S. suis 2 at 2 weeks after the 2 nd immunization. (B) Clinical symptom of leg and histopathology in the brain of immunized pigs challenged with SS2. (C) Western blots analysis the ability of sera from the pigs vaccinated with rSC0016 or rSC0018 containing pS-SaoA plasmid to recognize SaoA protein of SS2 and SS7. PB: Post-boost serum of pigs. (D) Rectal temperature of immunized piglets after challenged intravenously with 10 10 CFU of SS7. The rectal temperatures were measured at 0, 1, 2, 3, 4, 5, 6, and 7 days after challenge. The experiment was performed twice.