Outer membrane vesicles from genetically engineered Salmonella enterica serovar Typhimurium presenting Helicobacter pylori antigens UreB and CagA induce protection against Helicobacter pylori infection in mice

ABSTRACT

Helicobacter pylori causes globally prevalent infections that are highly related to chronic gastritis and even development of gastric carcinomas. With the increase of antibiotic resistance, scientists have begun to search for better vaccine design strategies to eradicate H. pylori colonization. However, while current strategies prefer to formulate vaccines with a single H. pylori antigen, their potential has not yet been fully realized. Outer membrane vesicles (OMVs) are a potential platform since they could deliver multiple antigens. In this study, we engineered three crucial H. pylori antigen proteins (UreB, CagA, and VacA) onto the surface of OMVs derived from Salmonella enterica serovar Typhimurium (S. Typhimurium) mutant strains using the hemoglobin protease (Hbp) autotransporter system. In various knockout strategies, we found that OMVs isolated from the ΔrfbP ΔfliC ΔfljB ΔompA mutants could cause distinct increases in immunoglobulin G (IgG) and A (IgA) levels and effectively trigger T helper 1- and 17-biased cellular immune responses, which perform a vital role in protecting against H. pylori. Next, OMVs derived from ΔrfbP ΔfliC ΔfljB ΔompA mutants were used as a vector to deliver different combinations of H. pylori antigens. The antibody and cytokine levels and challenge experiments in mice model indicated that co-delivering UreB and CagA could protect against H. pylori and antigen-specific T cell responses. In summary, OMVs derived from the S. Typhimurium ΔrfbP ΔfliC ΔfljB ΔompA mutant strain as the vector while importing H. pylori UreB and CagA as antigenic proteins using the Hbp autotransporter system would greatly benefit controlling H. pylori infection.

IMPORTANCE

Outer membrane vesicles (OMVs), as a novel antigen delivery platform, has been used in vaccine design for various pathogens and even tumors. Salmonella enterica serovar Typhimurium (S. Typhimurium), as a bacterium that is easy to engineer and has both adjuvant efficacy and immune stimulation capacity, has become the preferred bacterial vector for purifying OMVs after Escherichia coli. This study focuses on the design of Helicobacter pylori ;(H. pylori) vaccines, utilizing genetically modified Salmonella OMVs to present several major antigens of H. pylori, including UreB, VacA and CagA. The optimal Salmonella OMV delivery vector and antigen combinations are screened and identified, providing new ideas for the development of H. pylori vaccines and an integrated antigen delivery platform for other difficult to develop vaccines for bacteria, viruses, and even tumors.

Introduction

Helicobacter pylori is a Gram-negative bacterium that colonizes the human gastric mucosa and is the primary cause of the progression of gastric diseases, including chronic gastritis, peptic ulcers, and even gastric cancer, if not treated promptly [1]. Antibiotics are currently recognized as the preferred treatment for H. pylori infection in clinical practice. However, the alarming rate of bacterial antibiotic resistance has greatly overwhelmed antibiotic development [2]. Therefore, an increasing number of researchers are developing better treatment methods, with effective, safe, and low-cost vaccines against H. pylori infection being one feasible option [3]. The current design of H. pylori vaccines is mainly based on H. pylori live bacterial preparations, flagellin antigens, or adhesion antigens, none of which have met expectations. Consequently, strategies for designing novel H. pylori vaccines to eradicate H. pylori colonization remain a hot research topic.

Urease B (UreB), vacuolating cytotoxin A (VacA), and cytotoxin-associated antigen (CagA) are three crucial antigenic proteins in H. pylori that induce host immune responses [4]. Despite current vaccine formulations or antigen presentation strategies using these antigenic proteins providing strong immunogenicity and certain protective effects, they have shown defects such as low immunoprotective efficacy and insufficient specific immune responses in clinical trials. Therefore, exploring new antigen formulation strategies for H. pylori vaccines is a significant issue.

Outer membrane vesicles (OMVs) are extracellular vectors secreted by Gram-negative bacteria and always exist around them. They carry many pathogen-related molecular patterns to exert immune-stimulating effects [5,6]. Therefore, they have great potential in the design of vaccine vectors and immunologic adjuvants. Salmonella has joined Escherichia coli as the commonly engineered bacteria for developing vaccine delivery vectors due to their ease of modification by genetic engineering. However, to use Salmonella OMVs as vectors, the first urgent problem to address is the interference of autoantigens on their presented antigens while ensuring their presented antigens evoke an effective immune response in the body. Our research team previously discovered that truncated O-antigen (ΔrfbP) from Salmonella lipopolysaccharide effectively enhanced the cross-immune protection efficacy of OMVs. Simultaneous deletion of Salmonella flagellar structural proteins FliC and FljB promoted humoral and mucosal immunity to a certain extent and markedly improved cross-protection effects [7]. In addition, OMVs cross-protected against the same Salmonella serotype and heterologous E. coli after the knockout of outer membrane proteins (OMPs) OmpA, OmpC, and OmpD [8]. These results indicate that Salmonella OMVs can serve as an antigen delivery vector with great potential for developing broad-spectrum vaccines.

Therefore, this study aims to explore the effects of knocking out major OMPs (OmpA, OmpC, and OmpD) on antigen presentation and develop a strain with a stronger antigen presentation ability, ensuring the autoimmune protection of Salmonella OMVs and the capability of the presented protein antigens to be recognized by the host. Furthermore, the hemoglobin protease (Hbp) autotransporter system is used to express the primary H. pylori antigenic proteins (UreB, CagA, and VacA) on the surface of Salmonella OMVs, and the strength of immune protection and anti-H. pylori infection effects are analyzed in mice. Through these steps, we expect to select an optimal antigen combination to construct a candidate vaccine, which will undoubtedly pave the way for further prevention and treatment against H. pylori infection.

Results

Genetically engineered Salmonellaalmonella OMVs successfully present H. pylori antigen UreB

The autotransporter pathway, a branch of the type V secretion system, is the most common and typical mechanism Gram-negative bacteria use to secrete large virulence factors [9]. We used the Hbp autotransporter platform to replace the side domains d1 and d2 of the passenger domain with antigens UreB and VacA, respectively, and insert antigen CagA into the position of side domain d4, enabling the three antigens to be successfully expressed on the outer surface of Salmonella OMVs (Figure 1a) [10]. In addition, we adopted the suicide plasmid method to conduct knockout experiments with different combinations of OmpA, OmpC, and OmpD based on previous experiments on ΔrfbP, ΔfliC, and ΔfljB strains. Moreover, we introduced target antigenic proteins onto the surface of bacterial OMVs generated by different knockout strains with the Hbp autotransporter platform [7]. Taking UreB protein as the target, we conducted a western blot to validate the expression of the three antigenic proteins. Prominent bands were visible in the corresponding region of the protein from strains with UreB, indicating the successful import of the protein into the specific knockout bacteria (Figure 1b and Supplemental Figure S1), which was also confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Supplementary Figure S2).

Figure 1.

Construction, heterologous antigen protein characterization, and quantities of genetically engineered OMVs. (a) Using the Hbp autotransporter platform, the side domains d1 and d2 of the passenger domain were replaced by antigens UreB and VacA, respectively, and the CagA antigen was inserted into side domain d4. The three antigens were successfully expressed on the outer surface of Salmonella OMVs. (b) The western blot shows apparent bands of a size consistent with the UreB antigenic protein in all strains, indicating its successful introduction into the corresponding mutant strain. (c) The number of OMV particles per mL was normalized to the number of CFUs per mL for each mutant and parental strain.

To explore the effects of our knockout strategy on the number of OMVs released by the strains, we measured the OMV secretion levels in each group. Bacterial OMV production increased to some extent when the ompA gene was knocked out, with the highest OMV production achieved by knocking out the ompA gene while retaining the ompB/C genes (Figure 1c). In addition, particle size analysis of OMVs in each group showed that the diameters of bacterial OMVs were mainly distributed in the 60–100 nm range, regardless of whether the UreB protein was introduced (Supplementary Figure S3), indicating that the import of the UreB protein does not significantly impact the particle size of bacterial OMVs.

OmpA-deficient Salmonella OMVs presenting UreB evoke potential protection against H. pylori infection

We performed immunization-challenge experiments using strains of various knockout strategies. We used UreB as the target to compare the effects of H. pylori antigen presentation by OMVs from different Δomp strains with diverse antigenic proteins and clarify the induced systemic and mucosal immune responses. First, quantitative enzyme-linked immunosorbent assays (ELISAs) were used to determine the concentration of serum immunoglobulin G (IgG) in mice after immunization and the concentration of immunoglobulin A (IgA) in the gastric mucosa of mice in the eighth week. Two weeks after infection with significant doses of H. pylori by intragastric administration, we sacrificed the mice and collected their gastric tissues for urease tests and bacterial load determination (Figure 2a). All mice maintained good health during the immunization, and no abnormal behavior was observed. Regardless of the knockout strategy, the immune effect in induced mice was greater in the group treated with OMVs delivering UreB protein than in the groups treated with OMVs without the antigenic protein and phosphate-buffered saline (PBS). Therefore, the UreB antigenic protein introduced into OMVs can successfully trigger immune responses in mice, showing an excellent protective effect (Figures 2b-e).

Figure 2.

Serum IgG and IgA levels and protective efficacy of mice immunized with OMVs delivering UreB according to the protocol. (a) The immunization and challenge protocols for mouse experiments. Mice were separately immunized with parental OMVs on day 0 and boosted on day 30. Serum antibody levels were measured at weeks 2, 4, 6, and 8. Half the mice were sacrificed at week 8 to collect their spleen and stomach tissues to quantify cytokines and stomach sIgA. The H. pylori challenge test was performed at week 10, and all the remaining mice were sacrificed for urease tests and gastric bacterial load determination at week 12. All animal experiments were repeated twice, so data were collected from 18 mice for analysis, whether it was antibody testing that required the euthanizing of mice or routine non-invasive indicator testing on mice. Total serum amounts of anti-H. pylori UreB IgG (b) and IgA (c) in OMV-immunized mice. lgG and lgA levels were significantly higher in the OmpA group (p < 0.01). The highest IgG level was found with mutant strain QS0076 (ΔrfbP ΔfliC ΔfljB ΔompA with pHbp + UreB). The highest IgA level was found with mutant strain QS00102 (ΔrfbP ΔfliC ΔfljB ΔompA ΔompC ΔompD with pHbp + UreB). Urease (d) and bacterial load (e) tests showed that urease contents and bacterial loads were lower in mice immunized with OMVs from the ΔompA mutant strain than those not immunized with OMVs from the ΔompA mutant strain.

Additionally, serum IgG and gastric mucosal secretory IgA (S-IgA) levels in mice were significantly higher when immunized with OMVs secreted by strains with than without ompA knockout (p < 0.01). In addition, IgG levels were the highest in QS0076 (ΔrfbP ΔfliC ΔfljB ΔompA with pHbp + UreB), and IgA levels were the highest in QS00102 (ΔrfbP ΔfliC ΔfljB ΔompA ΔompC ΔompD with pHbp + UreB), indicating that OMVs produced by the ΔompA mutant strain were likely to evoke better humoral and mucosal immune responses (Figures 2b,c). The urease tests and bacterial load determinations also showed that the urease contents and bacterial loads of mice were generally lower when immunized with OMVs secreted by strains with than without ompA knockout, consistent with the IgG and IgA levels (Figures 2d,e). In summary, OMVs with H. pylori antigenic proteins effectively evoke protective immune responses. The ΔompA strain evokes the best immune protection effects among all knockout strategies and could be a candidate vaccine against H. pylori infection.

OmpA-deficient Salmonella OMVs presenting UreB mainly elicit Th1-biased immune responses

In mice, IgG₁ and IgG_2c levels are considered closely associated with T helper 1 (Th1) and 2 (Th2) responses, respectively. Clarifying the type of immune response induced by OMVs presenting H. pylori antigenic proteins provides information for further elucidating the immune mechanism and optimizing the immune effects. Therefore, we used quantitative ELISA to determine and compare serum IgG₁ and IgG_2c levels in mice to evaluate the bias of immune responses. IgG₁ and IgG_2c levels were notably higher in immunized mice than in control mice, demonstrating that OMVs delivering UreB protein could induce both Th1 and Th2 immune responses (Figures 3a,b). IgG₁ levels were remarkably higher in mice immunized with OMVs secreted by the ΔompA strain than the other evaluated strains, indicating that ΔompA could induce a more significant increase in IgG₁ levels (Figure 3a). Furthermore, the ratio of IgG₁/IgG_2c was close to 1 in QS00132 (ΔrfbP ΔfliC ΔfljB ΔompC ΔompD with pHbp + UreB), indicating nearly balanced immune responses. In contrast, the ratios in the other experimental groups were significantly greater than 1, with the increase greater for IgG₁ than for IgG_2c, suggesting that OMVs with UreB tended to predominantly elicit Th1-biased immune responses (Figure 3c).

Figure 3.

Specific antibody subclasses and relative ratio in mice immunized with OMVs delivering UreB determined by quantitative ELISA. The serum IgG₁ (a) and IgG_2c (b) levels in mice immunized with OMVs presenting the UreB protein. Each group comprised 18 mice. All animal experiments were repeated twice, so data was collected from 18 mice for analysis. The exact serum concentrations of IgG₁ and IgG_2c subclass antibodies in mice are shown at 4 and 8 weeks after immunization. (c) The ratio of IgG₁ and IgG_2c levels in mice was compared to determine Th1/Th2 immune response levels. While immune responses induced by OMVs from QS00132 tended to be balanced, those from the other mutant strains tended to induce Th1 immune responses. The means were compared using the one-way ANOVA analysis. A p < 0.05 was considered statistically significant.

OmpA-deficient Salmonella OMVs presenting UreB trigger apparent Th17 cellular immune responses

Immunizing mice with vaccines triggers strong humoral and mucosal immunity and different types of cellular immunity. Therefore, clarifying the type of cellular immunity induced by various OMVs in mice can further elucidate their effect and immune mechanism. Consequently, we examined the levels of cytokines produced by mouse splenocytes after immunization with OMVs secreted by strains with different knockout strategies presenting the UreB antigen to evaluate the effect of cellular immune responses they induced. We used ELISA to quantify the levels of the cytokines interleukin (IL)-17, IL-4, IL-12, and IL-6 produced by mouse mesenteric lymph node (MLN) cells and splenocytes after antigen stimulation. The IL-17 levels were markedly higher in the ΔompA-OMV-immunized group than in the other groups, indicating antigen-presenting OMVs secreted by OmpA-deficient strains effectively induced specific T helper 17 (Th17) cellular immune responses in mice (Figures 4a,e). This finding might relate to the enhanced invasion of gastric mucosa after ompA deletion, resulting in a more robust immune response.

Figure 4.

The levels of cytokines secreted by MLN cells and splenocytes. IL-17 (a and e), IL-4 (b and f), IL-12 (c and g), and IL-6 (h and i) levels secreted by MLN cells and splenocytes from mice immunized with OMVs presenting the UreB protein. MLN cells and splenocytes were isolated from mice after the final immunization and cultured with outer membrane proteins from H. pylori for 24 h. IL-17, IL-4, IL-12, and IL-6 levels in culture supernatants were determined using specific ELISAs. The cytokine production results are expressed as mean ± SD (n = 18). All animal experiments were repeated twice, so data was collected from 18 mice for analysis. The means were compared using the least-significant-difference test. A p < 0.05 or p < 0.01 was considered statistically significant.

IL-12 and IL-4 are critical cytokines in the Th1 and Th2 cellular immune responses, respectively [11]. The IL-12 and IL-4 levels were also significantly higher in all OMV-immunized groups presenting the UreB antigen than in the control group, indicating that OMV immunization triggered both Th1 and Th2 cellular immune responses in mice, consistent with previous results (Figures 4b, f, c, g). Notably, the IL-4 and IL-12 levels were generally higher in the group immunized with the ΔompA OMVs than in the groups immunized with the other knockout OMVs, indicating that ΔompA may be the most effective among the various knockout strategies. We measured the level of pro-inflammatory cytokine IL-6 in MLN cells and splenocytes in each knockout group. However, while IL-6 levels were generally increased, their increases were not as large as those of the other cytokines, laying the foundation for the preliminary safety evaluation of the immunization induced by our constructed OMV models (Figures 4d, h).

Identification of the optimal combination of genetically engineered Salmonella OMVs to present H. pylori antigens

The previous experiments have shown that the best comprehensive effects on humoral and cellular immunity were achieved with OMVs derived from the S. Typhimurium ΔrfbP ΔfliC ΔfljB ΔompA mutant strain coated with H. pylori UreB antigenic protein. H. pylori also has diverse antigenic proteins besides UreB, including VacA and CagA, that trigger the corresponding host immune responses [4]. To determine the optimal combination of the various H. pylori antigenic proteins, we introduced various antigenic protein (UreB, VacA, and CagA) combinations into the OMVs derived from the S. Typhimurium ΔrfbP ΔfliC ΔfljB ΔompA mutant strain. We used western blots to verify the successful expression of each antigenic protein on the surface of OMVs (Figure 5a and Supplemental Figure S4), which were supported by SDS-PAGE results (Supplemental Figure S5). Corresponding IgG and gastric mucosal S-IgA levels significantly increased in mice after immunization with different combinations of OMVs and H. pylori antigens (p < 0.01; Figures 5b, c). Antibody levels differed among groups immunized with different antigen combinations: anti-UreB IgG and IgA levels induced by QS0074 (ΔrfbP ΔfliC ΔfljB ΔompA with pHbp + UreB-CagA) were remarkably higher than those of the other groups (Figures 5b,c), demonstrating a more robust immune response to UreB when combined with CagA.

Figure 5.

Identification of H. pylori antigens on OMVs derived from the Salmonella mutant ΔrfbP ΔfliC ΔfljB ΔompA strain and corresponding IgG and IgA levels. (a) Different combinations of antigenic proteins UreB, VacA, and CagA were introduced into OMVs derived from the Salmonella mutant ΔrfbP ΔfliC ΔfljB ΔompA strain; western blotting confirmed that all antigenic proteins were successfully expressed on the surface of OMVs. Mice were immunized with OMVs with different antigen combinations. Their serum was collected four weeks after the final immunization to detect antigen-specific IgG levels (b), and their gastric tissue was collected after sacrifice to measure the antigen-specific IgA levels (c). All antibodies were measured by ELISA. The means were compared using the least significant difference test. A p < 0.01 was considered statistically significant.

Genetically engineered Salmonella OMVs presenting H. pylori antigens UreB and CagA protect against H. pylori and antigen-specific T cell responses

In order to evaluate the immune effects of different OMVs and select candidate vaccines, we immunized mice with diverse antigen combinations before infecting them with significant H. pylori doses by intragastric administration. The mice were sacrificed after two weeks, and their gastric tissues were collected for urease tests and bacterial load determinations to detect H. pylori infection. Consistent with the previous experiments, the urease contents and bacterial loads were notably lower among mice immunized with antigen-presenting OMVs than among negative control mice, indicating that the constructed OMVs offered satisfactory protection against H. pylori infection. The urease test results also showed that urease activity in gastric tissues was remarkably lower in mice immunized with QS0074 (ΔrfbP ΔfliC ΔfljB ΔompA with pHbp + UreB-CagA) than with the other antigen combinations (Figure 6a). The bacterial load results were consistent with the urease test results. Compared with other experimental strains, the QS0074 strain reduced the bacterial load in the stomach of mice (Figure 6b). The histopathological pictures of stomach tissues from various groups were scored and showed that mice immunized with OMVs derived from QS0074 strain indicated protection against H. pylori infection, and their gastric mucosa showed no abnormalities (Figure 6c). This finding demonstrated that OMVs secreted by the knockout strain delivering UreB and CagA antigens were a candidate vaccine to provide the adequate immune protection.

Figure 6.

Protective efficacy against H. pylori and antigen-specific CD4⁺ T cell immune response induced by OMVs derived from different antigen combination strategies. (a) Urease activity in gastric tissue homogenates from mice immunized with OMVs presenting various antigen combinations was assessed two weeks after the challenge infection. (b) H. pylori colony counts in gastric tissue homogenates were quantified to determine bacterial loading. The means were compared using the least significant difference test. A p < 0.05 or p < 0.01 was considered statistically significant. (c) Histopathological picture of gastric mucosa in mice immunized with various OMVs, and the corresponding pathology score of various groups were shown. (d-f) Antigen-specific CD4⁺ T cell immune responses. Eight weeks after the H. pylori challenge, the spleens were collected from mice to prepare a single-cell suspension. After staining with intracellular cytokines, flow cytometry was used to determine the proportion of CD4⁺ CD154⁺ T cells (d), CD154⁺ IFN-γ⁺ T cells (e), and CD154⁺ IL17A⁺ T cells (f) in total CD4⁺ T cells. The data are presented as mean ± SD. The means were compared using the least significant difference test.

To further evaluate the effects of the vaccine, we used flow cytometry to analyze and compare the levels of cluster of differentiation 4 (CD4)⁺ and 154 (CD154)⁺ T lymphocytes and CD4⁺ CD154⁺ T lymphocytes secreting interferon (IFN)-γ or IL-17A after stimulation with different antigen combinations in mice. The levels of mouse-specific CD4⁺ CD154⁺ T cells, CD4⁺ CD154⁺ IFN- γ⁺ T cells, and CD4⁺ CD154⁺ IL-17A⁺ T cells increased significantly after immunization with OMVs presenting H. pylori antigenic proteins, providing further evidence of the correctness of our strategy for vaccine design (Figures 6d-f). Moreover, the increase was higher with QS0074 than with the other strains, confirming that the protective effects against H. pylori infection were achieved by IFN-γ-mediated Th1 and IL-17A-mediated Th17 immune responses.

Materials and methods

Bacterial strains and growth conditions

All the strains used in this study are listed in Table 1. S. Typhimurium and related mutants were routinely cultured in Luria-Bertani broth or agar at 37°C for 36 ~ 48 h to the exponential phase before the experiment to ensure an adequate concentration and viability [14]. The H. pylori strains were cultured in 2.8% (w/v) Brucella broth base (Difco Labs, Detroit, MI, USA) supplemented with 5% fetal calf serum (Thermo Fisher Scientific, North Shore City, New Zealand) under microaerobic conditions (5% O₂, 10% CO₂, and 85% N₂) at 37°C with shaking at 120 rpm. Mouse-adapted H. pylori strain PMSS1 was a gift from Professor Yong Xie at the First Affiliated Hospital of Nanchang University in Nanchang, China. Suspensions of H. pylori strain were prepared from strains at fresh exponential phase to ensure a high level of viable cells for challenging.

Table 1.

Bacterial strains and plasmids used in this study.

Strain and plasmid	Description	Source
Strains
Helicobacter pylori
H. pylori 26695	The wild-type virulent strain	Lab collection
Escherichia coli
TOP10	Non-pathogenic E. coli engineering bacteria for plasmid construction	Lab collection
Salmonella enterica serovar Typhimurium (S. Typhimurium)
χ3761	S. Typhimurium UK-1	[12]
QS0071	S. Typhimurium UK-1 ∆rfbP ∆fliC ∆fljB ∆ompA	[8]
QS0072	ΔrfbP ΔfliC ΔfljB ΔompA with plasmid pHbp+UreB-VacA-CagA	This study
QS0073	ΔrfbP ΔfliC ΔfljB ΔompA with plasmid pHbp+UreB-VacA	This study
QS0074	ΔrfbP ΔfliC ΔfljB ΔompA with plasmid pHbp+UreB- CagA	This study
QS0075	ΔrfbP ΔfliC ΔfljB ΔompA with plasmid pHbp+VacA-CagA	This study
QS0076	ΔrfbP ΔfliC ΔfljB ΔompA with pHbp+UreB	This study
QS0077	ΔrfbP ΔfliC ΔfljB ΔompA with pHbp+VacA	This study
QS0078	ΔrfbP ΔfliC ΔfljB ΔompA with pHbp+CagA	This study
QS0081	ΔrfbP ΔfliC ΔfljB ΔompC	[8]
QS0082	ΔrfbP ΔfliC ΔfljB ΔompC with plasmid Hbp+UreB-VacA-CagA	This study
QS0091	ΔrfbP ΔfliC ΔfljB ΔompD	[8]
QS0092	ΔrfbP ΔfliC ΔfljB ΔompD with plasmid Hbp+UreB-VacA-CagA	This study
QS00101	ΔrfbP ΔfliC ΔfljB ΔompA ΔompC ΔompD	[8]
QS00102	ΔrfbP ΔfliC ΔfljB ΔompA ΔompC ΔompD with plasmid Hbp+UreB-VacA-CagA	This study
QS00111	ΔrfbP ΔfliC ΔfljB ΔompA ΔompC	[8]
QS00112	ΔrfbP ΔfliC ΔfljB ΔompA ΔompC ΔompD with plasmid Hbp+UreB-VacA-CagA	This study
QS00121	ΔrfbP ΔfliC ΔfljB ΔompA ΔompD	[8]
QS00122	ΔrfbP ΔfliC ΔfljB ΔompA ΔompD with plasmid Hbp+UreB-VacA-CagA	This study
QS00131	ΔrfbP ΔfliC ΔfljB ΔompC ΔompD	[8]
QS00132	ΔrfbP ΔfliC ΔfljB ΔompC ΔompD with plasmid Hbp+UreB-VacA-CagA	This study
Plasmid
pYA3337	Asd expression vector Ptrc promoter pSC101ori	[13]
pQS0003	For expression of Hbp + UreB-VacA-CagA antigen	This study
pQS0004	For expression of Hbp + UreB-VacA antigen	This study
pQS0005	For expression of Hbp + UreB-CagA antigen	This study
pQS0006	For expression of Hbp + VacA-CagA antigen	This study
pQS0007	For expression of Hbp + UreB antigen	This study
pQS0008	For expression of Hbp + VacA antigen	This study
pQS0009	For expression of Hbp + CagA antigen	This study

Construction of plasmids and related mutant strains

As mentioned above, homologous recombinant plasmids with knockouts of flagellin-related genes rfbP, fliC, and fljB and OMP-related genes ompA, ompC, and ompD were constructed by the suicide plasmid method and then transferred into S. Typhimurium χ3761 competent cells according to different combinations via electroporation. A series of expression-deficient Salmonella strains were obtained after monoclonal screening and colony PCR verification (QS0071, QS0081, QS0091, QS00101, QS00111, QS00121, and QS00131). The strains and their genotypes constructed and used in this experiment are shown in Table 1. Daleke-Schermerhorn et al. verified that the Hbp autotransporter platform could introduce multiple antigenic proteins into the vector plasmid and make them all express on the surface of OMVs [15]. Therefore, we adopted the genomic DNA of S. Typhimurium χ3761 as a template and used Hbp-F and Hbp-R as primers to amplify gene fragments of the Hbp autotransporter platform and the linearized plasmid pYA3337 simultaneously, and the corresponding primers are shown in Table 2. The Hbp autotransporter platform expression cassette of S. Typhimurium χ3761 was cloned into the pYA3337 vector using a Gibson assembly kit, which was then transferred into the E. coli strain TOP10 so that the constructed pYA3337-Hbp plasmid could be extensively amplified. We used the Hbp autotransporter platform to clone the antigenic protein gene sequence into the Hbp expression cassette with the recombinant plasmid pYA3337-Hbp as the vector.

Table 2.

The primers used in this study.

Primers	Sequences (5“→3”)	Function
UreB-F	5’TGCTGACAAAGAACTGGGTTCTACCGCAATTTTTG3’	Amplified UreB gene in H. pylori genome
UreB-R	5’GAAACAGTTCCTGAGTCGGCGTTCGATCACCCT3’
VacA-F	5’GCGTGAAACATAACGGTACCTTTATGTTGTTTGTG3’	Amplified VacA gene in H. pylori genome
VacA-R	5’CATATGTGGTGGTTCCCTGTGTTAATTGGTACCTGTAGAAAC3’
CagA-F	5’TGCCACCCTGAGTCTGAACAGCCTACTGATTACTTTGGTAAC3’	Amplified CagA gene in H. pylori genome
CagA-R	5’CGCTGTTACGACGCATTGAGACTTAAGATTTCTGGAAACCAC3’
Hbp-F	5’ACGCGTATATGCAAGTCCACCGGTTTAAG3’	Amplified the Hbp autotransporter system in S. Typhimurium
Hbp-R	5’TTCGAGCTGACTGACTGTAAGCGTACAGCCTG 3’
pYA3337-Hbp-UreB-F	5’AGGGTGATCGAACGCCATGACGTAACTGACGATTGC3’	Amplified the fragments of pYA3337-Hbp-UreB for construct the plasmid pQS0003, pQS0004, pQS0005 and pQS0007
pYA3337-Hbp-UreB-R	5’CAAAAATTGCGGTAGAACCAGTAAACCTTGGCGACGTG3’
pYA3337-Hbp-CagA-F	5’GTGGTTTCCAGAAATCTTAAAGCGCCGGCCTGAAGGTAATC3’	Amplified the fragments of pYA3337-Hbp-CagA for construct the plasmid pQS0003, pQS0005, pQS0006 and pQS0009
pYA3337-Hbp-CagA-R	5’GTTACCAAAGTAATCAGTGCGCCGAGAGCGATTGA3’
pYA3337-Hbp-VacA-F	5’GTTTCTACAGGTACCAATTCCGGCGCGAATAGCGATGACG3’	Amplified the fragments of pYA3337-Hbp-VacA for construct the plasmid pQS0003, pQS0004, pQS0006 and pQS0008
pYA3337-Hbp-VacA-R	5’CACAAACAACATAAAGGGCTAAATTACGACGAATTCTG3’

First, we adopted the H. pylori genome as a template and corresponding primers to amplify the UreB gene sequence. We used the one-step cloning technique with a Gibson assembly kit, which was subsequently cloned into the prepared recombinant plasmid pYA3337-Hbp. The product was then introduced into E. coli TOP10 cells to obtain the recombinant pYA3337-Hbp-UreB plasmid. Using primers corresponding to different antigenic protein gene sequences to amplify them in the H. pylori genome, we obtained recombinant plasmids pYA3337-Hbp-CagA and pYA3337-Hbp-VacA using the same approach.

Next, we adopted the H. pylori genome and vector plasmid pYA3337-Hbp-UreB as templates and corresponding primers to amplify the CagA antigenic protein gene and vector. After one-step cloning of the amplified product using a Gibson assembly kit, the product was transferred into E. coli TOP10 cells to obtain recombinant plasmid pYA3337-Hbp-UreB-CagA. The VacA gene sequence was introduced via the same method to obtain the recombinant pYA3337-Hbp-UreB-VacA plasmid. The CagA gene sequence was introduced into the vector pYA3337-Hbp-VacA to create the recombinant plasmid pYA3337-Hbp-VacA-CagA.

Similarly, we adopted the pYA3337-Hbp-UreB-CagA recombinant plasmid as the vector, which could be amplified using the upstream and downstream primers of pYA33327-Hbp-VacA. After one-step cloning of the amplified VacA gene sequence and the vector, the product was transferred into E. coli TOP10 cells to obtain the recombinant pYA3337-Hbp-UreB-CagA-VacA plasmid. Lastly, we purified the obtained recombinant plasmids and transferred them into S. Typhimurium competent cells to create defective Salmonella strains that stably express H. pylori antigenic proteins.

OMV purification, characteristics and nanoparticle tracking analysis

Native OMVs (without any detergent treatment) were collected and isolated from culture supernatants of S. Typhimurium χ3761 and its mutants as previously described [8,16]. Briefly, 2 L of bacteria culture supernatants in the logarithmic phase (OD₆₀₀ = 1) were collected and filtered using a 0.45 μm Steritop bottle-top filter unit (Millipore et al., USA). Next, the vesicles in the filtrate were precipitated by centrifugation (2 h 40,000 g, 4°C) and resuspended in Dulbecco’s PBS (DPBS; Mediatech, Manassas, VA, USA). Then, the vesicles were further purified via density gradient centrifugation (overnight, 200,000 g, 4°C) with a discontinuous OptiPrep density gradient medium (Sigma-Aldrich, St. Louis, MO, USA). Next, 2 mL of 20%, 25%, 30%, 35%, 40%, and 45% OptiPrep was added from top to bottom in 10 mM HEPES (pH = 6.8) with 0.85% NaCl. Then, the obtained vesicles were washed gently three times with DPBS and then dissolved in 1 mL DPBS. The total protein concentration was determined with the bicinchoninic acid assay (Thermo Fisher Scientific, USA) to determine OMV yield from the same cell mass of diverse mutants. Each OMV sample was analyzed by 12% SDS-PAGE, and the gels were stained using GelCode™ Blue stain reagent (Thermo Pierce). The ladder was used to indicate the molecular weight of proteins (Catalog Number 26,619, Thermo Pierce). All OMVs were purified and quantified three times. The self-made rabbit serum antibodies corresponding to the protein, including UreB, VacA, and CagA were used for immunoblotting to detect the expression of different proteins in OMVs. The OMV particle size distribution and concentration were determined by nanoparticle tracking analyses (NanoSight NS300, Malvern Instruments, United Kingdom) as previously described [8,17]. The average of triplicate video statistics was taken for each sample. In addition, the number of OMVs per mL was normalized to the colony-forming units (CFUs) per mL for each strain to determine the number of OMVs per CFU.

Ethics statement

All animal experiments were conducted according to the guidelines of ARRIVE guidelines and the Animal Welfare Act and relevant regulations of Nanchang University (Nanchang, China; approval no.: NCDXYD-2020-008). The Animal Welfare Committee of Nanchang University approved all animal work protocols. Every effort was made to minimize animal suffering during experiments.

Immunization and challenge schedule for animal experiments

The animals used in this experiment were female C57BL/6 mice (six weeks old, 16–22 g) purchased from the Laboratory Animal Science Center of Nanchang University. All mice underwent a one-week environmental adaptation period.

The first stage of animal experiments screened the optimal vector for antigenic proteins. The mice were divided into nine groups, with 18 per group. The experimental group comprised mice immunized with OMVs secreted by a strain with UreB, the blank control group comprised mice immunized with OMVs secreted by a strain without UreB (QS00101: ΔrfbP ΔfliC ΔfljB ΔompA ΔompC ΔompD), and the negative control group comprised mice gavaged with PBS buffer. After suspending 100 μg of purified OMVs in 200 μL of PBS, mice were immunized via intragastric administration; the negative control group was injected with 200 μL PBS buffer. A booster immunization was given on day 30 after the first immunization. On the day before and the second, fourth, sixth, and eighth week after immunization, the animals’ orbit was washed five times with 0.1 mL PBS, and blood was collected from their orbital venous plexus. The collected blood was centrifuged, and the serum was collected and stored at − 80°C for further experiments.

In addition, eight weeks after the first immunization, nine of the 18 mice in each group were sacrificed to obtain stomach and spleen tissues and MLN cells. The experiment was repeated twice, so data on 18 mice were collected for analysis. One-quarter of the stomach tissue was homogenized with 1 mL of PBS containing a protease inhibitor mixture and 0.05 M ethylenediamine tetraacetic acid. The homogenate supernatant was collected and diluted at 1:10 to determine specific gastric mucosal S-IgA levels. The spleen and MLNs were used to determine cytokine levels. The nine remaining mice were used for challenge experiments. The mice were orally challenged with 10⁸ CFU of the H. pylori PMSS1 strain in the tenth week after the first immunization. After two weeks of feeding, the mice underwent terminal anesthesia and cervical dislocations. Their gastric tissues were separated, cleaned, and used for urease tests and gastric bacterial load determinations.

The second phase of animal experiments was conducted to identify the combination of antigens that induces the best immune protective response. The mice were divided into nine groups of 18. The blank control group comprised mice immunized with OMVs secreted by the optimal vector QS0071 strain (ΔrfbP ΔfliC ΔfljB ΔompA), the experimental group comprised mice immunized with OMVs secreted by the QS0071 strain with different antigenic protein combinations, and the negative control group comprised mice injected with PBS buffer. Similarly, 100 μg of purified OMVs suspended in 200 μL of PBS was used to immunize mice in the experimental and blank control groups by intragastric administration, while mice in the negative control group were gavaged with 200 μL of PBS buffer. The follow-up experiment was conducted in the same manner as described above. Each stage of animal experiments was conducted twice, and the data were combined for analysis.

ELISA

The antibody levels were determined by quantitative ELISA. Specific experimental methods were consistent with former studies [18]. We isolated 1 μg of OMPs from H. pylori strain 26,695 and suspended them in 100 μL of sodium carbonate-bicarbonate coating buffer (pH = 9.6) per well in 96-well plates (Nalge Nunc International, Naperville, IL, USA) and incubated overnight at 4°C to coat the well. Next, purified mouse Ig isotype standards (IgG, IgG₁, IgG_2c, or IgA; BD Biosciences, MA, USA), which were diluted twice in advance, were used to construct standard curves for quantification of each antibody isotype. Biotinylated goat anti-mouse IgG, IgG₁, IgG_2c, and IgA were used as secondary antibodies to perform the quantitative ELISA. The appropriate standard curves were used to calculate the final Ig isotype concentration in the antibody samples. All ELISA experiments were conducted in triplicate.

Western blots

Antigenic proteins (UreB, VagA, CagA, and their combination) expressed by S. Typhimurium χ3761 mutants were verified by SDS-PAGE and western blotting. OMVs with or without antigen expression were denatured and separated by 12% SDS-PAGE and subsequently transferred onto nitrocellulose membranes, which were then blocked in PBS containing Tween 20 (PBST) and 5% skim milk at room temperature for 2 h to prevent nonspecific proteins binding. Next, the membranes were washed three times with PBST and incubated with properly diluted UreB-, VacA-, or CagA-specific monoclonal antibodies at 4°C overnight. The membranes were washed three times with PBST before being incubated with an alkaline phosphatase-conjugated goat anti-rabbit IgG secondary antibody (1:10,000 dilution; Sigma-Aldrich) at room temperature for 2 h. Finally, the membrane was washed three times with PBST before being detected using an HRP Western Blot Analysis Kit (BD Biosciences, MA, USA).

Detection of cytokine production

In order to evaluate the impact of vaccine injection on T cell (Th1/Th2) polarization in the body, splenocytes and MLN cells were isolated from mice to detect their IL-17, IL-4, IL-12, and IL-6 levels. Splenocytes and MLN cells were collected four weeks after the booster immunization with corresponding antigens and stimulated for 24 h with 6 μg/mL of OMPs isolated from H. pylori strain as previously reported [19]. The supernatants were collected from stimulated cells, and cytokine levels were detected by ELISA. Briefly, monoclonal anti-IL-17, anti-IL-4, anti-IL-12, and anti-IL-6 antibodies (BD Biosciences, CA, USA) were coated onto duplicate wells in 96-well plates overnight at 4°C and then blocked with PBS containing 1% bovine serum albumin. After washing, the wells were incubated with biotinylated monoclonal IL-17, IL-4, IL-12, and IL-6 antibodies (BD Biosciences, MA, USA). The reaction was started by adding 3,3,’5,5’-tetramethyl-benzidine (Moss Inc., Pasadena, CA, USA) and terminated by adding 0.5 M HCl. Standard curves were generated using recombinant mouse IL-17, IL-4, IL-12 and IL-6 proteins.

Urease test

A tissue strip was obtained from the stomach of each mouse and used to prepare an aliquot of gastric antrum tissue homogenate with 0.5 mL of 0.8% NaCl solution to detect the urease activity [20]. Tissue homogenates from mice injected with PBS or OMVs without H. pylori antigenic proteins served as the negative control. First, 3 mL of urea broth (1 mg/mL glucose, 1 mg/mL peptone, 2 mg/mL KH₂PO₄, 5 mg/mL NaCl, and 1% urea) containing the phenol red indicator was mixed with 100 μl of tissue homogenate and incubated at 37°C for 4 h. Then, Gastric urease activity was measured spectrophotometrically (OD = 550 nm).

Bacterial load determination

Two weeks after the challenge, the stomach tissues of all the mice were collected for bacterial quantification to assess the protection induced by immunization. The stomach was cut in half and washed with PBS. The tissues were placed in a pre-weighed 5 mL tube containing brain-heart infusion broth medium and weighed. The accuracy of the weighing instrument was 0.0001 grams. Tissue sections were immersed in sterile tissue homogenizers, diluted at 1:10, 1:100, and 1:1000, and inoculated on a sterile Campylobacter agar base (Difco) supplemented with 10% sheep blood under microaerobic conditions at 37°C for 6–7 days. H. pylori colonies were verified on the plates using the urease reaction, PCR technology, and mass spectrometry [18].

Histopathology

Mouse stomachs were obtained two weeks after H. pylori challenge. The pathological analysis is based on the previously reported method [21]. Briefly, stomach tissues were fixed in 10% neutral buffered formalin for at least 24 h. The trimmed tissues were processed, paraffin-embedded, sectioned at 5 mm, and stained with hematoxylin and eosin. The pathologist was blinded as to the experimental groups and treatments of the study. Histologic samples of both glandular and squamous portions of the stomach were examined and evaluated and scored based on intensity 0 to 6 for criteria including mucosal inflammation and type, submucosal inflammation and type, mucosal ulceration, and hyperkeratosis of squamous stomach.

Flow cytometry

Single splenocyte suspensions were supplemented with 100 IU/mL penicillin-streptomycin and 10% fetal bovine serum in 2 mL Roswell Park Memorial Institute medium in 12-well plates. Next, the cells were stimulated by incubating them with 2 μg of cluster of differentiation 28 (CD28) and 40 μg of OMVs secreted by different groups for 24 h at 37°C in 5% CO₂. In the last 6 h, the protein transport inhibitors brefeldin A (2 μL/mL; BD, NJ, USA) and monensin (1.4 μL/mL; BD) were added to the cells. Then, the supernatant was removed via centrifugation, and 2 μL of Fc receptor blocker (BioLegend, CA, USA) was added and incubated for 10 min at room temperature. Next, 5 μL of premixed True-Stain Monocyte Blocker (Biolegend) and 2 μL CD4 (Biolegend) were added and incubated at room temperature in the dark for 25 min. Then, the cells were washed with PBS, incubated in 1 mL of diluted Zombie Aqua solution (Biolegend) for 10 min, and centrifuged. Next, the cells were fixed, permeabilized, and washed according to the Fixation/Permeabilization Kit (BD) instructions. Every centrifugation could obtain a cell cluster that would then be added to premixed Brilliant Stain Buffer Plus (BD; 10 mL), True-Stain Monocyte Blocker (5 mL), anti-CD3 (Biolegend; 2 mL), anti-CD154 (Biolegend; 2 mL), anti-IL-17A (Biolegend; 2 mL), and anti-IFN-γ (Biolegend; 2 mL) and mixed thoroughly. The cells were incubated for 30 min at room temperature in the dark with BD Perm/GelCode™ Buffer and PBS before being resuspended in 0.2 mL PBS. The number of various cell types was measured on the flow cytometer.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 5 software (GraphPad Software, San Diego, CA, USA). The experimental and control groups were compared using a one-way analysis of variance. Data are expressed as means ± standard deviation (SD). A p < 0.05 was considered statistically significant.

Discussion

Wild type H. pylori infection generally does not elicit immune protective response from the host, making it difficult to eradicate H. pylori colonization through natural infection induced immune response. This is also one of the reasons why most patients experience chronic H. pylori infection. H. pylori infection is the primary cause of gastritis, peptic ulcers, and even gastric cancer, making it of great practical significance to explore novel preventive and therapeutic H. pylori vaccines. However, previous studies preferred to prepare vaccines by combining a single H. pylori antigen with various adjuvants. For example, Zhang et al. adopted H. pylori antigen UreB as the immunogen and combined it with IL-2 as the adjuvant [22]. Similarly, Liu et al. used polysaccharide antigens as adjuvants to induce anti-H. pylori infection effects [23]. Unfortunately, while these vaccine preparation strategies stimulated the human body to produce specific antibodies and reduced the H. pylori load and urease activity in the stomach, the immune responses and effects to H. pylori colonization remained inadequate, limiting their practical efficacy.

In this study, we simultaneously imported three important H. pylori antigenic proteins (UreB, VacA, and CagA) onto the surface of OMVs from mutant Salmonella strains in different combinations to prepare vaccines and then identified the optimal strategy: simultaneously introducing UreB and CagA into the S. Typhimurium ΔrfbP ΔfliC ΔfljB ΔompA mutant strain successfully elicited strong humoral and mucosal immunity in mice, and markedly increased IgG and IgA levels. IFN-γ and IL-17 have been proven to be vital in immunity against H. pylori infection [24], and CD154 expression on the surface of activated CD4⁺ T cells determines the degree of differentiation of CD4⁺ T cells into Th1 and Th17 cells after immunization [25]. Our data showed that the constructed vaccine induced significant increases in CD4⁺ T cells relevant to preventing H. pylori infection and related cytokines IFN-γ and IL-17 levels. Declines in H. pylori colonization and urease activity in mice were apparent after H. pylori challenge experiments. All these results suggest that our novel vaccine strategy has promising potential in preventing and treating H. pylori infection (Figure 7).

Figure 7.

Summary diagram of the mechanisms by which engineered OMVs present the H. pylori antigen proteins UreB, CagA, and VacA as vaccines.

We adopted the Hbp system to express antigenic proteins on the surface of OMVs, which stimulated immune responses in the body more effectively than intravesicular antigens [26]. The Hbp system, a crucial mechanism through which Gram-negative bacteria perform secretory functions, helps transport proteins across their outer membrane and is an indispensable mechanism for exporting virulence and adhesion factors [27]. This system comprises three parts: an N-terminal signal peptide, a C-terminal β-barrel domain, and a middle passenger domain. Studies have shown that replacing the passenger side domains with antigenic proteins did not affect the basic functions of the Hbp system and enabled antigenic proteins to be expressed on the surface of OMVs. For example, Jong et al. successfully expressed Mycobacterium tuberculosis antigenic proteins ESAT6, Ag85B, and Rv2660c in E. coli strains with the Hbp system [10]. Similarly, Bart et al. used the Hbp system to express SpyTag genes on the surface of Salmonella, further demonstrating the efficacy of the Hbp system in effectively expressing heterologous proteins on the surface of OMVs [28]. In this study, we fully used the characteristics of the Hbp autotransporter system, importing three H. pylori antigenic protein genes and obtaining OMVs expressing multiple heterologous antigenic proteins simultaneously (Figure 1 and Supplemental Figure S1).

Bacterial OMVs include periplasmic proteins, OMPs, lipopolysaccharides, and other components of pathogen-related molecular patterns with stronger immunogenicity than single antigens. Their bimolecular lipid layer structures prevent antigenic proteins from rapid clearance [29]. Our previous studies have shown that ΔrfbP caused the remodeling of the outer membrane structure, increasing OMV production and conducive to obtaining a more appropriate immune reactivity intensity [30]. Moreover, the ΔfliC ΔfljB mutants effectively reduced the contamination of OMVs with flagellin to induce better cross-immunity. Our preliminary study demonstrated the potential of using these engineered Salmonella OMVs to deliver Shigella flexneri 2a O-antigen polysaccharides as a vaccine delivery platform [8,31]. Therefore, diverse OMP knockout strategies were performed in ΔrfbP ΔfliC ΔfljB mutant strains with different effects in inducing immune responses, with the ΔompA strains inducing significantly stronger immune responses and anti-infection effects than the other mutant strains. This finding might be because OmpA could impact OMV biosynthesis by interacting with other lipoproteins (e.g. Lpp and NlpA) to affect the level of antigens exposed on the OMV surface, leading to a decline in the immune effects of vaccines in the body [32,33]. Additionally, the altered effects of antigenic-protein-induced immune responses were likely caused by spatial structure interactions between OMPs or between OMPs and heterologous antigenic proteins.

It is well-known that cytokines and related lymphocytes are crucial in the body’s immune system. For example, IFN-γ and IL-12, mainly secreted by Th1 cells, are closely associated with cellular immunity [34,35]; IL-4, secreted by Th2 cells, contributes significantly to the activation and maintenance of humoral immunity [36]; IL-6 and IL-17, mainly secreted by Th17 cells, play a role in host defense against extracellular pathogens, and IL-6 is used as a marker of mucosal inflammation to evaluate the safety of OMV vaccines [37]. CD4⁺ T cells are currently thought to play a major role in the fight against H. pylori infection, and various subtypes secrete various cytokines to participate in body immunity. The findings of previous studies indicate that Salmonella OMVs could effectively stimulate Th1 immune responses in mice, while Th17 immune responses were not evident [38].

In this study, after the intragastric administration of Salmonella OMVs delivering UreB antigenic protein to mice, their serum IgG₁ level was significantly higher than their IgG_2c level, and the IL-17 level in spleen lymphocytes and MLN cells increased markedly after stimulation with H. pylori antigens. These results showed that the vaccine induced Th1-biased and Th17 immune responses simultaneously, which might be due to the UreB antigen presented by Salmonella OMVs binding to pattern-recognition receptors (e.g. Toll-like receptors) on gastric mucosal epithelial cells to activate intracellular signaling pathways, thereby promoting the secretion of abundant cytokines and then activating Th17 cells to produce IL-17. Among the cytokines produced after stimulation of H. pylori antigens, the low IL-6 level foreshadowed the satisfactory safety of our vaccine (Figure 4). A previous study found that the differentiation of Th17 cells was mostly associated with the IL-6/signal transducer and activator of transcription 3 (STAT3) signaling pathway [37], while low IL-6 levels and significantly elevated IL-17 levels suggested that IL-17 secretion could be induced by some other potential pathway after H. pylori antigen presentation by Salmonella OMVs [39]. Interestingly, IL-17 showed a greater elevation than IL-4, IL-12, and IL-6 in this experiment (Figure 4), likely due to its involvement as a pro-inflammatory factor in regulating local inflammation induced by H. pylori infection. Since IL-17 may have multiple effects, the level of immune protection and inflammatory responses must be considered comprehensively in the subsequent vaccine development to achieve a balance and develop an optimal vaccine design strategy.

There remain some areas for improvement. For example, the immune effects of other H. pylori antigenic proteins imported into Salmonella OMVs should be examined in the future. In addition, the location of different antigenic proteins in the Hbp autotransporter system should be optimized to trigger immune responses more effectively. Moreover, the strain chosen for the challenge experiment is the mouse-adapted H. pylori strain PMSS1. Considering the genetic variability of H. pylori and the limitations of current animal model (not all clinical isolates can infect mice), we will continue to use the Mongolian gerbil animal model and the gerbil-adapted H. pylori 7.13 strain to evaluate the protective effect of the vaccine against H. pylori infection in further research. If given the opportunity, even non-human primate models will be attempted to evaluate vaccine potential [40,41]. Since we aim to utilize the H. pylori infection model to preliminarily assess the best combination of antigens with the OMV vector, the adopted model was infected with H. pylori for two weeks to evaluate the protective effects. Therefore, pathological changes were not evaluated. However, it is necessary to conduct some pathological analyses (e.g. tissue section analysis) to ensure the vaccine’s protective effects and cytotoxicity assessments to evaluate its immunogenicity and safety for clinical application.

In conclusion, our study demonstrated the promising potential of OMVs from mutant Salmonella strains with multiple H. pylori antigenic proteins expressed on their surface as a vaccine. Among the diverse examined strategies, immunizing mice with OMVs from the S. Typhimurium ΔrfbP ΔfliC ΔfljB ΔompA mutant strain as the vector with UreB and CagA of H. pylori as antigenic proteins may significantly increase humoral and mucosal immunity. Our findings provide innovative measures and data to support further H. pylori vaccine development. In addition, this novel S. Typhimurium OMVs vector can not only delivering H. pylori antigens, but also serve as a vector platform for delivering antigens or other protein drugs for treating diseases such as antibiotic-resistant bacteria, viral infections, and even tumors, which is expected to play a great clinical application potential.

Funding Statement

This study was supported by the National Natural Science Foundation of China [82203032, 32260193 and 32060040], Natural Science Foundation of Jiangxi Province [20202BAB206062], Training Plan for Academic and Technical Leaders of Major Disciplines in Jiangxi Province-Youth Talent Project [20212BCJ23036], Project for high and talent of Science and Technology Innovation in Jiangxi “Double-Thousand Talents Program of Jiangxi Province” [jxsq2023301110 and jxsq2023201019] and the National Innovation and Entrepreneurship Training Program for college students [202110403093 and 202310403042].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

Qiong Liu and Xiaotian Huang conceived and designed the experiments; Qiong Liu, Yinpan Shang, Lu Shen, Xiaomin Yu, Yanli Cao, Lingbing Zeng, Hanchi Zhang, ZiRong Rao, and Yi Li performed the experiments. Ziwei Tao and Zhili Liu analyzed the data; Qiong Liu, Yinpan Shang, Lu Shen, and Xiaotian Huang wrote the manuscript.

Compliance with ethical standards

All animal experiments were conducted according to the guidelines of ARRIVE guidelines and the Animal Welfare Act and relevant regulations of Nanchang University (Nanchang, China; approval no.: NCDXYD-2020-008).

Data availability statement

The data that support the findings of this study are openly available in figshare at http://10.6084/m9.figshare.25670886

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2024.2367783