Application of an O-Linked Glycosylation System in Yersinia enterocolitica Serotype O:9 to Generate a New Candidate Vaccine against Brucella abortus

Brucellosis is a major zoonotic public health threat worldwide, causing veterinary morbidity and major economic losses in endemic regions. However, no efficacious brucellosis vaccine is yet available, and live attenuated vaccines commonly used in animals can cause human infection. N- and O-linked glycosylation systems have been successfully developed and exploited for the production of successful bioconjugate vaccines. Here, we applied an O-linked glycosylation system to a low-pathogenicity bacterium, Yersinia enterocolitica serotype O:9 (Y. enterocolitica O:9), which has repeating units of O-antigen polysaccharide (OPS) identical to that of Brucella abortus (B. abortus), to develop a bioconjugate vaccine against Brucella. The glycoprotein we produced was recognized by both anti-B. abortus and anti-Y. enterocolitica O:9 monoclonal antibodies. Three doses of bioconjugate vaccine-elicited B. abortus OPS-specific serum IgG in mice, significantly reducing bacterial loads in the spleen following infection with the B. abortus hypovirulent smooth strain A19. This candidate vaccine mitigated B. abortus infection and prevented severe tissue damage, thereby protecting against lethal challenge with A19. Overall, the results indicated that the bioconjugate vaccine elicited a strong immune response and provided significant protection against brucellosis. The described vaccine preparation strategy is safe and avoids large-scale culture of the highly pathogenic B. abortus.


Introduction
Brucellosis is one of the most common zoonotic diseases worldwide and it has re-emerged in several countries in recent years [1]. More than 500,000 new cases of brucellosis are reported each year [2], and millions of livestock are either infected or at risk [3]. Incidence is likely significantly underestimated because of the wide distribution, mild clinical features, and frequent misdiagnosis of the disease [4]. Brucellosis is mainly caused by Brucella spp., which are Gram-negative and facultative intracellular α-proteobacteria. Brucella melitensis, Brucella abortus, and Brucella suis are the most virulent of these species. Humans are mainly infected by direct contact with infected animals or through the consumption of non-pasteurized dairy products [5]. In humans, infection is chronic and patients gradually lose their ability to work. Although rarely fatal, human brucellosis is very serious [6,7], requiring long-term antibiotic treatment and often causing permanent sequelae [8].
At present, animal vaccines against Brucella are live attenuated vaccines, and these have played an important role in controlling the spread of brucellosis [9]. However, the efficacy of these vaccines is not ideal [10]-they can cause abortions and infertility in animals, disease in humans [11,12], and

Glycoprotein Purification
IPTG-induced cells (7 L) were collected by centrifugation and then resuspended in 300 mL of Buffer A1 (20 mM Tris-HCl, pH 7.5, 10 mM imidazole, and 0.5 M NaCl), after which the cells were disrupted by ultrasound for 2 h. The supernatant of the cell lysate was harvested by centrifugation at 10,000 ×g for 20 min and then loaded onto a nickel affinity column (cOmplete His-Tag Purification Resin, Roche, penzberg, Germany) that had been pre-equilibrated with five column volumes of Buffer A1. The flow rate was 4 mL/min. After washing with ten column volumes of Buffer A1, the target glycoprotein was eluted in Buffer A2 (20 mM Tris-HCl, pH 7.5, 0.5 M imidazole, and 0.5 M NaCl). The eluent was concentrated with a 10 kDa cut-off centrifugal filter (Merck, Darmstadt, Germany) to less than 10 mL. The concentrated solution was separated through a Superdex 200 Prep Grade column (16 mm × 1000 mm; GE Healthcare, Beijing, China) with phosphate-buffered saline (PBS) at a flow rate of 1 mL/min. The fractions were collected and analyzed by 12% SDS-PAGE.

Lipopolysaccharide (LPS) and OPS Extraction
LPS extraction was performed as described previously [35]. Briefly, the culture was collected by centrifugation and then resuspended in ddH 2 O. After freezing and thawing, an equal volume of 90% phenol was added to each sample followed by vigorous shaking at 68 • C and then centrifugation to collect the uppermost part of the supernatant (the water layer). The phenol layer was re-extracted with ddH 2 O and the previous step was repeated. The extracts were dialyzed into ddH 2 O, then DNase (5 µg/mL; Solarbio, Beijing, China), RNase (1 µg/mL; Solarbio), and proteinase K (20 µg/mL; Merck) were sequentially added to the dialyzed samples and incubated at the optimal temperature for each enzyme. After a boiling water bath for 10 min, samples were centrifuged to obtain the LPS.
Glacial acetic acid was added to the extracted LPS to achieve a final concentration of 1% (v/v). After boiling for 90 min, the pH of the LPS extract was adjusted to 7.0 with NaOH. Finally, the mixture was centrifuged at 40,000 ×g for 5 h, and the supernatant was collected as the OPS fraction.

Animal Immunization Experiments
Specific-pathogen-free female BALB/c mice were purchased from Charles River and housed in the Laboratory Animal Centre of the Academy of Military Medical Sciences at constant temperature and humidity. Water, food, and bedding were changed every 4 days. All animal experiments were approved by and conducted in accordance with the recommendations of the Animal Care and Use Committee of the Academy of Military Medical Sciences (ethics approval code IACUC-DWZX-2017-023, approved in November 2017).
Groups of 10 5-week-old BALB/c mice were used in the immunization experiments. Purified glycoproteins or OPS were diluted with PBS, and aluminum hydroxide adjuvant (General Chemical Corp, Brighton, MI, USA) was added to 10% of the final volume. Mice were immunized intraperitoneally (i.p.) with three doses (2.5 µg of the polysaccharides/mouse) at 2-week intervals and blood samples were collected by tail snip on the 10th day following the third dose. Mouse sera were separated from the blood and stored at −20 • C. Two weeks after the final immunization, mice were injected i.p. with different doses of A19 as described below, followed by determination of the bacterial load in the mouse spleens and monitoring of mouse survival.

Enzyme-Linked Immunosorbent Assay (ELISA)
A 96-well immunoplate was precoated with 10 µg/mL poly-L-lysine (100 µL per well). ELISA was performed as described previously [35]. Briefly, the 96-well immunoplate was coated for 2 h with diluted LPS at 37 • C and then washed three times with Wash Buffer (PBS + 0.05% Tween 20). The plates were patted dry and Blocking Buffer (PBS + 5% milk powder) was added to each well followed by incubation at 37 • C for 2 h. After drying the plates, diluted serum from the immunized mice was added to each well and the plate was incubated at 37 • C for 1 h. After another washing and drying step, 1:15,000 diluted HRP-conjugated goat anti-mouse IgG antibody (Abcam, Shanghai, China) was added to each well and incubated at 37 • C for 1 h. The washing and drying step was repeated. The Soluble TMB Kit (CWbio, Beijing, China) was used to initiate the detection reaction. Stop solution (2 M H 2 SO 4 ) was added to each well to stop the reaction, and the absorbance of each well was measured at a wavelength of 490 nm with a microplate reader.

Determination of Bacterial Loads in Spleens
Mice were immunized and a non-lethal dose of B. abortus A19 was injected i.p. on the 14th day after the third immunization. On the 7th day post-infection, mice were sacrificed by neck dissection, and the spleens were removed and placed in 1.5 mL microcentrifuge tubes. The spleens were weighed and then homogenized with normal saline. After centrifugation at 5000 ×g for 10 min, the supernatant was discarded and the pellet was resuspended with normal saline. After repeating the previous step twice, the pellet was resuspended with 1 mL of 0.1% (w/v) sodium deoxycholate. Finally, the bacterial suspension was diluted with normal saline and cultured on solid TSB medium. Bacterial colonies were counted after 2 days of culture at 37 • C.

Determination of Cytokine Levels
The cytokine levels of the immunized mice were determined with the Mouse TNF-α Precoated ELISA Kit, Mouse IL-1β Precoated ELISA Kit, Mouse IL-2 Precoated ELISA Kit, and Mouse IL-6 Precoated ELISA Kit (Dakewe, Shenzhen, China) in accordance with the manufacturer's instructions. Briefly, diluted serum samples and standards were added to the pre-coated wells, followed by the addition to each well of the biotinylated antibody relevant to the kit being used and incubation at 37 • C for 90 min. The wells were washed four times with Washing Buffer and patted dry, after which, Streptavidin-HRP was added to each well and the plates were incubated at 37 • C for 30 min. After another washing and drying step, TMB was added to each well and incubated at 37 • C for 15 min shielded from the light, and the reactions were terminated using Stop solution. The absorbance of each well was measured at a wavelength of 450 nm with a microplate reader.

Hematoxylin and Eosin (HE) Staining
The livers and spleens of immunized mice were fixed with 4% paraformaldehyde (Solarbio, Beijing, China), and then paraffin-sectioned and stained using the Hematoxylin-Eosin (HE) Staining Kit (Solarbio) in accordance with the manufacturer's instructions. Briefly, the fixed samples were embedded and sectioned, then conventional dewaxing and hydration were performed. The tissue sections were stained with HE, after which, dehydration, cleaning for transparency, and neutral resin sealing were performed successively.

Statistical Analysis
Antibody titers were log 2 -transformed and bacterial loads were log 10 -transformed. Statistical analyses were conducted using GraphPad Prism version 7.0 (GraphPad, San Diego, CA, USA).

Application of the O-linked Glycosylation System in Y. Enterocolitica O:9
Studies have reported that most polysaccharides are recognized by PglL [34], and our previous works demonstrated the successful use of PglL in some Gram-negative bacteria [35,38]. To create the conjugate vaccine against B. abortus with the O-linked glycosylation system, we first needed to verify that the OPS of B. abortus could be recognized by PglL. Y. enterocolitica O:9 strain 52212 (YeO9_52212) was used as the host cell because this low-pathogenicity strain has an identical OPS structure to that of B. abortus [37]. The pET-pglL-CTB4573H plasmid, from which PglL and rCTB [35] were co-expressed, or the pET-CTB4573H plasmid, from which rCTB alone was expressed, were introduced into YeO9_52212. After induction with IPTG and overnight culture, the total protein was extracted and separated by SDS-PAGE. Coomassie Blue staining and western blotting showed that the molecular weight (MW) of rCTB when expressed alone was only around 15 kDa ( Figure 1A). The MW of rCTB increased when PglL was co-expressed, indicating that rCTB might have been glycosylated. Further, two bands for rCTB were observed in YeO9_52212 cells expressing pET-pglL-CTB4573H. The lower band showed a slightly shifted MW compared with CTB4573H and the higher band had a MW between 35-40 kDa. These results indicated that the O-linked glycosylation system was efficiently expressed in YeO9_52212 and that almost all of the substrate protein had been glycosylated.
Glycoproteins were isolated by Ni 2+ affinity and size exclusion chromatography ( Figure S1). The recovery efficiency following purification was approximately 150 µg polysaccharide/L medium. Anti-B. abortus and anti-Y. enterocolitica O:9 specific monoclonal antibodies were used to test the structural similarity of OPS between Y. enterocolitica O:9 and B. abortus. The higher MW band of glycosylated CTB4573H (CTB-OPS Ba ) was detected by both antibodies, while the lower band was not observed because of the poor immunogenicity of short-chain OPS ( Figure 1B). CTB naturally exists in a pentameric form [39]. To verify whether glycosylated CTB4573H was also in a polymeric state, we performed native gel electrophoresis. Coomassie Blue staining showed that the MW of CTB-OPS Ba (C-OPS Ba ) was around 242 kDa ( Figure 1C), which is approximately 5-fold that of the C-OPS Ba monomer (~40 kDa). Thus, rCTB could still pentamerize after glycosylation and purification. Glycoproteins were isolated by Ni 2+ affinity and size exclusion chromatography ( Figure S1). The recovery efficiency following purification was approximately 150 μg polysaccharide/L medium. Anti-B. abortus and anti-Y. enterocolitica O:9 specific monoclonal antibodies were used to test the structural similarity of OPS between Y. enterocolitica O:9 and B. abortus. The higher MW band of glycosylated CTB4573H (CTB-OPSBa) was detected by both antibodies, while the lower band was not observed because of the poor immunogenicity of short-chain OPS ( Figure 1B). CTB naturally exists in a pentameric form [39]. To verify whether glycosylated CTB4573H was also in a polymeric state, we performed native gel electrophoresis. Coomassie Blue staining showed that the MW of CTB-OPSBa (C-OPSBa) was around 242 kDa ( Figure 1C), which is approximately 5-fold that of the C-OPSBa monomer (~40 kDa). Thus, rCTB could still pentamerize after glycosylation and purification.

Induction of Specific Antibody Responses in Mice
After confirming the conservation of C-OPS Ba between Y. enterocolitica O:9 and B. abortus, we next assessed the immunogenicity of C-OPS Ba through a series of animal experiments. After a preliminary assessment of the safety of this vaccine obtained by measuring biochemical indicators including ALP, AST, ALT, and BUN ( Figure S2), we tested the serum titers of antibodies against YeO9_52212 LPS following C-OPS Ba immunization.
Ten 5-week-old BALB/c mice were immunized i.p. with either purified C-OPS Ba or C-OPS Ba adjuvanted with aluminum hydroxide (C-OPS Ba +Al) on Days 0, 14, and 28. Each group was immunized with the same quantity of polysaccharide (2.5 µg/mouse) and another group was immunized with PBS only as a control. On the 10th day following the third immunization, serum was collected by tail snip and used to characterize IgG responses against YeO9_52212 LPS by ELISA. C-OPS Ba was observed to induce antibody titers against YeO9_52212 LPS in mice ( Figure S3). We further detected specific antibodies against B. abortus LPS.
Purified C-OPS Ba and OPS Ba from YeO9_52212, with or without 10% aluminum hydroxide adjuvant (C-OPS Ba , C-OPS Ba +Al, OPS Ba , OPS Ba +Al), were used for immunization as described above. ELISA results showed that B. abortus LPS-specific IgG titers in all groups were elevated compared with the PBS-vaccinated group, especially in C-OPS Ba -and C-OPS Ba +Al-vaccinated mice. The titer in the C-OPS Ba group was significantly higher than the titers of the OPS Ba and OPS Ba +Al groups but similar to that of the C-OPS Ba +Al group (Figure 2A). We further detected IgG subclass titers (IgG1, IgG2a, IgG2b, and IgG3) against B. abortus A19 LPS in the sera of PBS-, OPS Ba -, and C-OPS Ba -vaccinated mice. Apart from the IgG3 titers, other subclass titers in the C-OPS Ba group were significantly higher than in the PBS group ( Figure 2B).
including ALP, AST, ALT, and BUN ( Figure S2), we tested the serum titers of antibodies against YeO9_52212 LPS following C-OPSBa immunization.
Ten 5-week-old BALB/c mice were immunized i.p. with either purified C-OPSBa or C-OPSBa adjuvanted with aluminum hydroxide (C-OPSBa+Al) on Days 0, 14, and 28. Each group was immunized with the same quantity of polysaccharide (2.5 μg/mouse) and another group was immunized with PBS only as a control. On the 10th day following the third immunization, serum was collected by tail snip and used to characterize IgG responses against YeO9_52212 LPS by ELISA. C-OPSBa was observed to induce antibody titers against YeO9_52212 LPS in mice ( Figure S3). We further detected specific antibodies against B. abortus LPS.
Purified C-OPSBa and OPSBa from YeO9_52212, with or without 10% aluminum hydroxide adjuvant (C-OPSBa, C-OPSBa+Al, OPSBa, OPSBa+Al), were used for immunization as described above. ELISA results showed that B. abortus LPS-specific IgG titers in all groups were elevated compared with the PBS-vaccinated group, especially in C-OPSBa-and C-OPSBa+Al-vaccinated mice. The titer in the C-OPSBa group was significantly higher than the titers of the OPSBa and OPSBa+Al groups but similar to that of the C-OPSBa +Al group (Figure 2A). We further detected IgG subclass titers (IgG1, IgG2a, IgG2b, and IgG3) against B. abortus A19 LPS in the sera of PBS-, OPSBa-, and C-OPSBavaccinated mice. Apart from the IgG3 titers, other subclass titers in the C-OPSBa group were significantly higher than in the PBS group ( Figure 2B). (B) IgG subclass titers (IgG1, IgG2a, IgG2b, and IgG3) against A19 LPS were measured in sera from PBS-, OPSBa-, and C-OPSBa-vaccinated mice. Each value represents the mean ± standard deviation of log2-transformed titers in the sera of individual mice (shown as data points on the graphs) from each group (n = 10). The unpaired t-test was used to evaluate differences between IgG titers (**, p <0.01; ****, p <0.0001; ns, no statistically significant difference).

Evaluation of Vaccine-Induced Protection in Mice following Infection with a Non-Lethal dose of B. Abortus
Given the success of the bioconjugate vaccine in eliciting both Y. enterocolitica O:9 and B. abortus LPS-specific IgG antibodies, we further evaluated the protective effects of this vaccine against nonlethal infection. For traditional live attenuated vaccines against B. abortus, the virulent smooth strain of B. abortus has to be used to evaluate the protective effects (mainly through measurement of the clearance rate of bacteria in the mouse spleen). In contrast, there is no unified evaluation standard for (B) IgG subclass titers (IgG1, IgG2a, IgG2b, and IgG3) against A19 LPS were measured in sera from PBS-, OPS Ba -, and C-OPS Ba -vaccinated mice. Each value represents the mean ± standard deviation of log 2 -transformed titers in the sera of individual mice (shown as data points on the graphs) from each group (n = 10). The unpaired t-test was used to evaluate differences between IgG titers (**, p <0.01; ****, p <0.0001; ns, no statistically significant difference).

Evaluation of Vaccine-Induced Protection in Mice following Infection with a Non-Lethal dose of B. Abortus
Given the success of the bioconjugate vaccine in eliciting both Y. enterocolitica O:9 and B. abortus LPS-specific IgG antibodies, we further evaluated the protective effects of this vaccine against non-lethal infection. For traditional live attenuated vaccines against B. abortus, the virulent smooth strain of B. abortus has to be used to evaluate the protective effects (mainly through measurement of the clearance rate of bacteria in the mouse spleen). In contrast, there is no unified evaluation standard for the bioconjugate type of vaccines, and preliminary evaluation of their protective effect can be evaluated through infection with the B. abortus hypovirulent smooth strain. Fourteen days after the third immunization, mice were challenged with a non-lethal dose (1.03 × 10 7 CFU/mouse) of B. abortus hypovirulent strain A19 (A19). For all treatment groups, the TNF-α levels began to decline after reaching a peak on the 5th day ( Figure 3A). Serum TNF-α levels in the PBS-immunized group and the OPS Ba -and OPS Ba +Al-immunized groups increased markedly following infection. In contrast, C-OPS Ba and C-OPS Ba +Al vaccination significantly inhibited the production of TNF-α, and TNF-α levels in the C-OPS Ba group were lower than those in the C-OPS Ba +Al group. We also measured serum IL-1β, IL-2, and IL-6 levels in the immunized mice but no significant differences in expression were detected. . Immune responses of mice following non-lethal B. abortus A19 infection. Immunized mice were infected intraperitoneally with 1.03 × 10 7 CFU of A19 on the 14th day following the third immunization. As a control, another group of naive mice was injected intraperitoneally with normal saline. (A) After infection, the sera of mice in each group were collected on the 1st, 3rd, 5th, and 7th day and the TNF-α levels were measured. The unpaired t-test was used to evaluate differences between TNF-α levels at different time points. Each value represents the mean ± standard deviation Figure 3. Immune responses of mice following non-lethal B. abortus A19 infection. Immunized mice were infected intraperitoneally with 1.03 × 10 7 CFU of A19 on the 14th day following the third immunization. As a control, another group of naive mice was injected intraperitoneally with normal saline. (A) After infection, the sera of mice in each group were collected on the 1st, 3rd, 5th, and 7th day and the TNF-α levels were measured. The unpaired t-test was used to evaluate differences between TNF-α levels at different time points. Each value represents the mean ± standard deviation (n = 3). (B) On the 7th day post-infection, mouse spleens were collected and weighed and the bacterial loads were measured. Each value represents the mean ± standard deviation of spleen weight or log 10 -transformed bacterial loads (CFU/spleen) of individual mice (shown as data points on the graphs) from each group (n = 5 per group). The unpaired t-test was used to evaluate differences between spleen weights or bacterial loads (***, p <0.001; ****, p <0.0001; ns, no statistically significant difference). (C) The livers and spleens of infected mice and normal mice (Control) were fixed with 4% paraformaldehyde, paraffin sectioned, and then stained with hematoxylin and eosin. The yellow boxes in the top panels represent the field of view in the corresponding figures below, which were magnified four times. Green arrows indicate multinucleated giant cells and blue arrows indicate hyperplastic nodules.
Seven days following infection, the mice were dissected and their spleens were removed. We found that the spleen size had increased following infection, with a significant increase in the spleen weight in the PBS group compared with the C-OPS Ba group, and in the OPS Ba group compared with the control group ( Figure S4 and Figure 3B). We enumerated the bacteria in each spleen and observed approximately 10 3 -10 3.5 CFU of A19 in C-OPS Ba and C-OPS Ba +Al-vaccinated mice. In contrast, the bacterial loads of other groups remained around 10 4 -10 4.5 CFU in the spleen, which was significantly higher than those of mice immunized with C-OPS Ba and C-OPS Ba +Al ( Figure 3B). These results indicated that mice immunized with C-OPS Ba and C-OPS Ba +Al were effectively protected against invading pathogens and the resulting lethal cytokine storms.
Spleens and livers were paraffin-sectioned and stained with HE. Compared with the control mice, the immunized mice showed different degrees of pathological changes in their spleens and livers. The spleens of all vaccinated mice with low-spectroscopy observation showed an increase in the white pulp, reflecting the proliferation of lymphocytes. Multinucleated giant cells with varying numbers and shapes of nuclei caused by foreign bodies increased with high-spectroscopy observation (green arrows) ( Figure 3C). The livers showed an aggregation of lymphocytes under low-spectroscopy observation and hyperplastic nodules caused by Brucella were seen under high-spectroscopy observation (blue arrows) ( Figure 3C). Spleen and liver damage were most severe in mice immunized with PBS, followed by OPS Ba and OPS Ba +Al. This damage was mildest in the C-OPS Ba and C-OPS Ba +Al groups, especially the C-OPS Ba group ( Figure 3C).

Evaluation of Protection against the Lethal dose Challenge
After demonstrating the safety and efficacy of the vaccine against non-lethal infection, we next evaluated its protective efficacy against lethal challenge. Separate groups of mice were immunized with PBS, OPS Ba , OPS Ba +Al, C-OPS Ba , or C-OPS Ba +Al as described previously and challenged with a lethal dose of A19 ( Figure 4A). Mice were challenged i.p. with approximately 3 × LD 50 of A19 (1.54 × 10 8 CFU/mouse) 2 weeks after the last vaccination, and survival was monitored for up to 14 days. All of the PBS-vaccinated mice died within 2 days and the 2-week survival rates in the OPS Ba and OPS Ba +Al groups were no more than 60%. In contrast, vaccination with C-OPS Ba or C-OPS Ba +Al provided 100% protection without the adjuvant ( Figure 4B). These results demonstrated that the administration of C-OPS Ba was far more effective than OPS Ba in protecting mice against lethal challenge with B. abortus. These findings were consistent with previous studies.
To verify the stability of the protective effect of C-OPS Ba in mice, we immunized another three groups of mice with PBS, C-OPS Ba or C-OPS Ba +Al as described above to assess their survival rate following a higher lethal dose challenge (2.51 × 10 8 CFU/mouse, 5 × LD 50 ). With this higher challenge dose, the survival rate of the mice was 90% for the C-OPS Ba -and C-OPS Ba +Al-vaccinated mice ( Figure 4C), indicating that C-OPS Ba can induce potent protection in mice. The IgG titers against A19 LPS in the sera of these mice were then measured. There was a significant difference in titer between the C-OPS Ba -and PBS-vaccinated mice ( Figure S5A). Finally, immunized mice were infected with an even higher dose of A19 (lethal to 20% of PBS-vaccinated mice). The C-OPS Ba -and C-OPS Ba +Al-vaccinated mice showed significantly lower bacterial loads in the spleen compared with PBS-vaccinated mice ( Figure S5B). These results indicated that C-OPS Ba can induce stable protection in mice. To verify the stability of the protective effect of C-OPSBa in mice, we immunized another three groups of mice with PBS, C-OPSBa or C-OPSBa+Al as described above to assess their survival rate following a higher lethal dose challenge (2.51 × 10 8 CFU/mouse, 5 × LD50). With this higher challenge dose, the survival rate of the mice was 90% for the C-OPSBa-and C-OPSBa+Al-vaccinated mice ( Figure  4C), indicating that C-OPSBa can induce potent protection in mice. The IgG titers against A19 LPS in the sera of these mice were then measured. There was a significant difference in titer between the C-OPSBa-and PBS-vaccinated mice ( Figure S5A). Finally, immunized mice were infected with an even higher dose of A19 (lethal to 20% of PBS-vaccinated mice). The C-OPSBa-and C-OPSBa+Al-vaccinated mice showed significantly lower bacterial loads in the spleen compared with PBS-vaccinated mice ( Figure S5B). These results indicated that C-OPSBa can induce stable protection in mice.

Discussion
In this study, we developed a novel bioconjugate vaccine candidate against B. abortus based on a bacterial O-linked glycosylation system. YeO9_52212 was engineered to act as the host organism for the production of the bioconjugate vaccine against B. abortus. Because YeO9_52212 is less pathogenic and more easily cultured than B. abortus, we were able to avoid the large-scale culture of highly pathogenic bacteria. Our results demonstrated that C-OPSBa induced strong immune responses against B. abortus even in the absence of the aluminum hydroxide adjuvant.
We chose OPS as the target for vaccine development because of its ability to elicit antibodies that can confer protection against infection [40]. The glycoprotein generated in YeO9_52212 was able to induce antibodies against B. abortus as a result of the similarity in OPS between these two bacterial species [37]. The OPS of Brucella spp. is a homopolymer of N-formyl-perosamine (4-formamido-4,6dideoxy D-mannose) with various proportions of α-(1→2) and α-(1→3)-linkages and it has three antibody reactivities (A, M, and C). The A epitope corresponds to five or more continuous sugar units with α-(1→2) linkages and is typical of some B. abortus and B. suis biovars. The M epitope corresponds to the pentasaccharide with four α-(1→2)-linked and one α-(1→3)-linked polymers of the same sugar and is characteristic of B. melitensis. The C epitope represents the ratio of A/M reactivities highly

Discussion
In this study, we developed a novel bioconjugate vaccine candidate against B. abortus based on a bacterial O-linked glycosylation system. YeO9_52212 was engineered to act as the host organism for the production of the bioconjugate vaccine against B. abortus. Because YeO9_52212 is less pathogenic and more easily cultured than B. abortus, we were able to avoid the large-scale culture of highly pathogenic bacteria. Our results demonstrated that C-OPS Ba induced strong immune responses against B. abortus even in the absence of the aluminum hydroxide adjuvant.
We chose OPS as the target for vaccine development because of its ability to elicit antibodies that can confer protection against infection [40]. The glycoprotein generated in YeO9_52212 was able to induce antibodies against B. abortus as a result of the similarity in OPS between these two bacterial species [37]. The OPS of Brucella spp. is a homopolymer of N-formyl-perosamine (4-formamido-4,6-dideoxy D-mannose) with various proportions of α-(1→2) and α-(1→3)-linkages and it has three antibody reactivities (A, M, and C). The A epitope corresponds to five or more continuous sugar units with α-(1→2) linkages and is typical of some B. abortus and B. suis biovars. The M epitope corresponds to the pentasaccharide with four α-(1→2)-linked and one α-(1→3)-linked polymers of the same sugar and is characteristic of B. melitensis. The C epitope represents the ratio of A/M reactivities highly repeated in the OPS of all biovars [2,23]. B. abortus contains~98% of the A epitope, B. suis has a unique 1:7 ratio of α-(1→3)to α-(1→2)-linked polymers, and B. melitensis has only the M epitope of the pentasaccharide repeating unit [2]. Y. enterocolitica O:9 generates strong cross-reactivity with Brucella as its OPS is a homopolymer of N-formyl-perosamine with exclusively α-(1→2)-linkages [2,22]. It is thus reasonable to speculate that the bioconjugate vaccine produced in YeO9_52212 would induce cross-reactive immune responses against B. suis and B. melitensis. Cross-protection against B. suis or B. melitensis remains to be verified, and this could further expand the scope of vaccine application.
Although the use of AcrA as a carrier protein failed to exert a protective effect in mice against B. abortus infection in a previous report, the feasibility of the use of carriers proteins as part of the B. abortus vaccine preparation strategy was demonstrated [22]. Since carrier proteins play an important role in immune response, protein toxins such as the tetanus toxoid, the recombinant exotoxin A protein from Pseudomonas aeruginosa (rEPA), the diphtheria toxoid, and rCTB used in the present study, are usually included in the vaccine design for better stimulation of the immune system [24]. Our previous studies found that different carrier proteins have different immune effects, and we demonstrated that the use of rCTB in bioconjugate vaccines conferred superior protection against Shigella flexneri and Salmonella enterica serovar Paratyphi A than rEPA, another commonly used carrier protein [35,38]. The bioconjugate vaccine in the present study used rCTB and exhibited a potent immune effect that was probably attributable to the good immunoadjuvant properties of CTB itself [39], as CTB has been shown to induce anti-inflammatory responses and regulate T-cell immune responses [41]. In addition, the rCTB pentamer can form more complex spatial structures, which facilitates activation of the immune response. We found that protein glycosylation did not change the pentameric structure of the substrate rCTB protein, as shown by the native PAGE results in this study. This finding also suggested that the size, structure, charge, and other physical/chemical properties of carrier proteins should be considered for the rational design of vaccines in the future.
Because there is no clear and unified evaluation model for Brucella vaccines, we used TNF-α as an indicator of the protective effect of the vaccine. TNF-α is one of the most important proinflammatory cytokines, and its secretion is associated with inflammatory responses caused by infection. TNF-α is one of the first cytokines released by macrophages following B. abortus infection, and its production results from direct interactions between Brucella and macrophages [42]. TNF-α plays a crucial role in enhancing IL-12 production [23], and depletion of IL-12 leads to decreased production of interferon-γ and nitric oxide, resulting in exacerbation of infection [23,42]. Here, the serum TNF-α levels in mice immunized with C-OPS Ba were significantly lower than those in the other treatment groups after infection with B. abortus. Infection via the intraperitoneal route can induce a strong inflammatory response. There was a higher level of recall response in the C-OPS Ba -vaccinated group because of pre-existing protective antibodies, which could quickly neutralize and kill the invading pathogens, such that the bacterial loads in the spleen were significantly reduced. Because the pathogen could be rapidly cleared, there was no strong immune response; in contrast, the mouse spleens were enlarged in the PBS-and OPS Ba -immunized groups, with a significantly increased proportion of white pulp observed.
The two most common routes of human infection for Brucellosis are the ingestion of contaminated dairy products or inhalation of contaminated aerosols. This leads to Brucella colonization of organs such as the spleen, liver, and lymph nodes. Owing to its potential spread by aerosols, Brucella has been classified as a biological threat [43]. This poses a risk during the production of live attenuated vaccines, requiring stricter production requirements and resulting in higher production costs. Thus, the live attenuated vaccines could lead to the possibility of leakage contamination. Research and development of new candidate vaccines are urgently required. Our study provides a conceptual advance for vaccine production against B. abortus-not only is the described production process safer and more economical than that used for traditional live attenuated vaccines, but our strategy also avoids the potential risk of virulence reversion and antibiotic resistance. Our conjugate vaccine could also be used to diagnose brucellosis [22]. In addition, subunit vaccines make it easy to distinguish between immunization and natural infection in animals. To further improve safety, in future studies we will delete the virulence-related genes of the YeO9_52212 engineered bacteria to render it even more suitable for industrial production.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-2607/8/3/436/s1, Figure S1: Purification of CTB-OPS Ba . Figure S2: Biochemical indicators in the sera of mice after three immunizations. Figure S3: IgG responses against YeO9_52212 LPS. Figure S4: Spleen sizes of control (normal) and immunized mice following infection with B. abortus A19. Figure S5: IgG responses against B. abortus A19 LPS and bacterial loads in the spleens of PBS-, C-OPS Ba -, and C-OPS Ba +Al-vaccinated mice following infection with a higher dose of B. abortus A19.