Introduction

Acinetobacter baumannii can cause severe and fatal infections, including pneumonia, wound infections, and bloodstream infections, in immunocompromised patients, with mortality rates as high as 75% in untreated cases1,2. The persistence and rapid dissemination of A. baumannii in hospital environments are primarily attributed to its resistance to disinfectants, broad resistance to antimicrobial agents, and production of diverse virulence factors, which facilitate prolonged colonization and transmissions1,3. Therefore, the US Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) consider it crucial to research new approaches against A. baumannii4,5. Treatment of multidrug-resistant (MDR) A. baumannii infection in patients admitted to ICUs presents substantial challenges because the bacterium can quickly spread to critical organs, resulting in systemic infections associated with high mortality rates6. The increasing mortality rate associated with MDR A. baumannii sepsis and the inadequate efficacy of existing antimicrobial agents highlights the urgent need to develop effective vaccines to combat A. baumannii infections. Despite promising preclinical results from subunit vaccine candidates, including outer membrane, fimbrial, and biofilm-associated proteins, no authorized vaccine against A. baumannii infections has been approved7,8,9,10. This may be attributed to the need to incorporate adjuvants or delivery systems into their formulations to enhance stability and immunogenicity11,12.

Many A. baumannii strains isolated from patients express the highly conserved outer membrane protein 22 (Omp22), which possesses the outer membrane protein A (OmpA)-like domain, facilitates interactions with peptidoglycan and regulates the formation of outer membrane vesicles (OMVs), serving as carriers for the secretion of cytotoxic, virulence-enhancing, or resistance-inducing molecules7,13. Furthermore, owing to its limited similarity to human proteins, antibodies against Omp22 do not cross-react with human proteins7. Previous studies of Omp22 vaccine candidates revealed that its immunogenicity is closely linked to its delivery method and the adjuvants usage14,15,16. For instance, immunization of mice with 50 µg of Trx-Omp22 fusion protein with alum adjuvant protected against A. baumannii sepsis, whereas lower doses were less effective and showed cytotoxic effects on mammalian cell growth7. Another study demonstrated that immunization with 40 µg of Omp22 encapsulated in chitosan-poly (lactic-co-glycolic acid) increased survival rates against A. baumannii challenge compared to Omp22 alone14. Engineered OMVs displaying Omp22 also conferred dose-dependent protection against A. baumannii16. These findings highlight the importance of combining Omp22 with adjuvants or protein carriers to enhance its effectiveness and influence the immune response against A. baumannii.

Evidence indicates that flagellin, functioning as a pathogen-associated molecular pattern (PAMP), enhances the immunogenicity and protective efficacy of antigens through the activation of Toll-like receptor 5 (TLR5), which is expressed across diverse immune cell types17,18,19. Flagellin’s potent adjuvant activity is strictly linked to its high-affinity binding to TLR5 on CD11c+ dendritic cells (DCs), enhancing antigen uptake and presentation, thus leading to antigen-specific CD4+/CD8+ T cells and humoral immunity against pathogens17,18. Utilization of the flagellin D3 hypervariable region for integrating foreign antigens has been an effective strategy for stimulating both innate and adaptive immune responses through improved antigen presentation and TLR5-dependent mechanisms20,21. Integrating an envelope protein from the West Nile virus into the D3 domain of flagellin led to the development of a vaccine capable of eliciting neutralizing and protective antibodies against this virus21. Similarly, incorporating the hemagglutinin sequence of the avian influenza virus or Yersinia pestis V antigens into the hypervariable region of flagellin proved effective in developing vaccines against respective infections22,23. Accordingly, modification of flagellin constructs, such as replacing the variable region of flagellin D3 with the antigen, enhances immunogenicity, and triggers specific immune responses, underscoring the versatility and effectiveness of flagellin-based vaccines.

Previous studies have indicated that the immunogenicity of Omp22 can be considerably enhanced by employing various adjuvants and display systems14,15,16. In light of these findings, the current study aimed to investigate the potential of integrating Omp22 into the hypervariable D3 domain of flagellin (FliC-Omp22) as a vaccine candidate to enhance protective humoral and cellular immune responses against A. baumannii in a murine model.

Results

Design and characterization of FliC-Omp22 protein

The construction and design of FliC-Omp22 involved deleting the D3 domain (residues 185 to 285) of the native FliC protein and inserting Omp22 in its place (Fig. 1A). The FliC-Omp22 fusion protein exhibited a half-life of > 20 h, as determined by Expasy’s ProtParam analysis. Predictions indicated the stability of the chimeric FliC-Omp22 protein, with an instability index (II) of 28.04, classifying it as a stable protein. Computational modeling and prediction identified the surface-exposed loops of Omp22 spanning positions 103–113, 114–142, and 143–149 within the FliC-Omp22 fusion protein as linear and discontinuous B-cell epitopes (Fig. 1B). A 3D structural model of the FliC-Omp22 fusion protein was generated using the Robetta server (Fig. 1C). The FliC-Omp22 chimeric vaccine sequence had a VaxiJen score of 0.8431, indicating its potential as an antigen. SDS-PAGE revealed bands corresponding to approximately 25 kDa for Omp22, 51 kDa for FliC, and 64 kDa for FliC-Omp22 recombinant proteins (Supplementary Fig. S4). Additionally, recombinant proteins were identified in their purified form using a monoclonal anti-His-tag antibody (Fig. 1D).

Fig. 1
figure 1

Design and evaluation of fusion FliC-Omp22 protein. (A) Diagram illustrating the design and construction of FliC-Omp22. The D3 domain of FliC protein (residues 185 to 285) was removed and substituted with the Omp22 protein. (B) Schematic representation of the linear B-cell epitopes identified by BepiPred-1.0, BepiPred-2.0, ABCpred, LBtope, BcePred and BCPred tools. (C) Robetta’s prediction for the 3D structure of FliC-Omp22. (D) Western blot analysis shows purified recombinant Omp22, FliC, and FliC-Omp22 proteins in lanes 1, 2, and 3, respectively, identified using an anti-His tag antibody.

Full size image

Antigenicity, structural integrity and bioactivity of FliC-Omp22 fusion protein

To validate the appropriate display of the Omp22 protein and retention of FliC protein antigenicity in the FliC-Omp22 fusion protein, an ELISA assay was conducted using anti-Omp22 or anti-FliC specific antibodies. As illustrated in Fig. 2A, the FliC-Omp22 fusion protein effectively bound to both anti-Omp22 and anti-FliC-specific antibodies, indicating the accurate presentation of the Omp22 domain in the fusion protein and preservation of antigenicity for both Omp22 and FliC. Immunoblot analysis revealed that both anti-Omp22 and anti-FliC-specific antibodies specifically interacted with a ∼64 kDa protein, corresponding to the FliC-Omp22 protein (Fig. 2B and C). Additionally, the anti-FliC antibody displayed specific reactivity with ∼51 kDa and ∼64 kDa proteins, corresponding to FliC and FliC-Omp22 proteins, respectively (Fig. 2B). Similarly, the anti-Omp22 antibody showed specific reactivity with ∼25 kDa and ∼64 kDa proteins, corresponding to Omp22 and FliC-Omp22 proteins, respectively (Fig. 2B). Moreover, the anti-FliC-Omp22 antibody exhibited reactivity with ∼25 kDa, ∼51 kDa, and ∼64 kDa proteins, corresponding to Omp22, FliC, and FliC-Omp22 proteins, respectively (Fig. 2D). Additionally, anti-FliC and anti-Omp22 specific antibody did not show reactivity with the Omp22 and FliC proteins, respectively (Fig. 2B and C). To evaluate the TLR5-specific signaling potency of purified FliC-Omp22, TLR5-negative, and TLR5-positive RAW 264.7 cells were incubated with increasing concentrations of FliC-Omp22 or FliC (as a positive control). As illustrated in Fig. 2E, FliC-Omp22, similar to FliC, induced TNF-α production in a TLR5-specific manner and exhibited dose-dependent behavior. No significant difference was detected in the capacity of varying concentrations of FliC-Omp22 and FliC proteins to stimulate TNF-α production in TLR5-positive RAW 264.7 cells (P > 0.05; Fig. 2E). The effectiveness of both proteins was comparable, implying that inclusion of the Omp22 sequence into FliC did not substantially influence the TLR5 binding domain of flagellin. In contrast, Omp22 and PBS did not stimulate TNF-α production in TLR5-positive RAW 264.7 cells. No statistically significant differences were detected among FliC-Omp22, FliC, Omp22, and PBS in inducing TNF-α production in TLR5-negative RAW 264.7 cells (Supplementary Fig. S2).

Fig. 2
figure 2

Antigenicity, integrity and bioactivity of FliC-Omp22 fusion protein. (A) In the protein ELISA assay, recombinant FliC-Omp22 showed binding to anti-Omp22 and anti-FliC specific antibodies, compared to normal mouse serum, which showed no binding. (B) In the immunoblot analysis, anti-FliC antibody specifically reacted with FliC (lane 2, ∼51 kDa) and FliC-Omp22 (lane 3, ∼64 kDa) proteins and did not show reactivity with Omp22 (lane 1). (C) Anti-Omp22 antibody specifically bound to Omp22 (lane 1, ∼25 kDa) and FliC-Omp22 (lane 2, ∼64 kDa) proteins and did not show reactivity with FliC proteins (lane 3). (D) Anti-FliC-Omp22 antibody exhibited reactivity with FliC (lane 1, ∼51 kDa), Omp22 (lane 2, ∼25 kDa), and FliC-Omp22 (lane 3, ∼64 kDa) proteins. Lane M denotes the positions of molecular weight markers. (E) FliC-Omp22, similar to FliC, induced TNF-α production in TLR5-positive RAW 264.7 cells in a concentration-dependent manner. The data are represented as the mean ± standard deviation (SD).

Full size image

FliC-Omp22 immunization enhanced humoral immune responses

To evaluate the humoral responses elicited by the FliC-Omp22 vaccine, we conducted a comparative analysis of serum total IgG, IgG1 and IgG2asubtypes levels in immunized mice against whole-cell A. baumannii, Omp22, FliC, or FliC-Omp22 recombinant proteins (Fig. 3A-B). FliC-Omp22-immunized mice exhibited significantly higher serum IgG levels against A. baumannii compared to other groups (P < 0.05; Fig. 3A). Omp22-immunized mice also had higher anti-A. baumannii IgG levels than those receiving FliC or PBS (P < 0.05).

Fig. 3
figure 3

Antibody responses in in the sera of both immunized and non-immunized. (A) Total IgG levels were measured against Omp22, FliC-Omp22, FliC, and the whole cell of A. baumannii. (B) IgG1 and IgG2a subclass titers were assessed. The data are represented as the mean ± SD. Datasets marked with distinct superscript letters exhibit statistical differences from each other.

Full size image

Mice immunized with FliC-Omp22 displayed significantly higher serum IgG levels against the Omp22 protein on days 14, 28, and 42 post-immunization compared to other groups (P < 0.05; Fig. 3A). Following booster doses, these IgG levels in sera FliC-Omp22-immunized mice of further increased by 1- to 2-fold at four and six weeks post-booster, surpassing those in Omp22-immunized mice (P < 0.05; Fig. 3A). IgG subclass analysis revealed that both IgG1 and IgG2a levels against the Omp22 protein were markedly elevated in FliC-Omp22-immunized mice, with IgG1 levels predominating (P < 0.05; Fig. 3B). Similar patterns were observed in Omp22-immunized mice, where both IgG1 and IgG2a levels were higher than those in FliC or PBS-treated groups (P < 0.05; Fig. 3B).

Mice immunized with FliC-Omp22 exhibited significantly higher serum IgG levels against the FliC-Omp22 recombinant protein on days 14, 28, and 42 post-immunization compared to other groups (P < 0.05; Fig. 3A). These IgG levels continued to rise significantly four and six weeks post-booster, exceeding those in FliC or Omp22-immunized mice (P < 0.05). IgG subclass analysis indicated that both IgG1 and IgG2a levels against the FliC-Omp22 protein were markedly elevated in FliC-Omp22-immunized mice, were significantly elevated, with IgG1 consistently higher than IgG2a (P < 0.05; Fig. 3B).

Mice immunized with FliC showed the highest serum IgG levels against the FliC protein on days 14, 28, and 42 post-immunization, significantly surpassing those of other groups (P < 0.05; Fig. 3A). Six weeks post-booster, these IgG levels increased two-fold compared to FliC-Omp22-immunized mice (P < 0.05; Fig. 3A). IgG subclass analysis revealed significantly higher levels of both IgG1 and IgG2a in FliC-immunized mice, with IgG1 predominating (P < 0.05). Although FliC-Omp22-immunized mice also exhibited elevated IgG1 and IgG2a levels, they were lower than those in FliC-immunized mice (P < 0.05; Fig. 3B).

FliC-Omp22 immunization induced robust splenocyte proliferation and cytokine secretion

The effects of FliC-Omp22 immunization on splenocyte proliferation and cytokine production were assessed by evaluating the proliferation stimulation index (SI) and levels of IL-4 and IFN-γ following stimulation with Omp22, FliC, or FliC-Omp22 proteins. Splenocytes from FliC-Omp22-immunized mice showed significantly higher proliferation when stimulated with Omp22, with the stimulation index (SI) reaching up to 2.7 times that of splenocytes from Omp22-immunized mice (P < 0.01; Fig. 4A). This increased proliferation was accompanied by elevated IL-4 and IFN-γ production, with cytokine levels up to twice those observed in Omp22-immunized groups (P < 0.01; Fig. 4B), indicating a strong immune response to Omp22 in FliC-Omp22-immunized mice.

Fig. 4
figure 4

In vitro immune response of splenocytes from immunized and non-immunized mice. Two weeks post the final immunization, spleens from mice were processed and stimulated with Omp22, FliC-Omp22, and FliC antigens in vitro for 72 h. (A) The proliferative assay of splenocytes was measured with the MTT method. (B) The IL-4 and IFN-γ cytokine production in splenocytes stimulated by antigens were assessed. The data are represented as the mean ± SD. Datasets marked with distinct superscript letters exhibit statistical differences from each other.

Full size image

Stimulation with FliC-Omp22 led to a significantly higher SI in splenocytes from FliC-Omp22-immunized mice, up to 2.4 times greater than that of Omp22 or FliC immune splenocytes (P < 0.01; Fig. 4A). Cytokine production was also markedly increased, with IL-4 and IFN-γ levels up to twice those in splenocytes from Omp22 or PBS-immunized mice (P < 0.01; Fig. 4B).

Splenocytes from FliC-immunized mice exhibited the highest proliferation rate when stimulated with FliC, with the SI peaking at three times that of FliC-Omp22-immunized splenocytes (P < 0.05; Fig. 4A). Similarly, cytokine production was significantly elevated, with IL-4 and IFN-γ levels up to twice those observed in splenocytes from Omp22 or PBS-treated mice (P < 0.01; Fig. 4B). In both FliC and FliC-Omp22-immunized splenocytes, IL-4 levels surpassed IFN-γ levels (P < 0.05; Fig. 4B), indicating a strong Th2 response to FliC stimulation.

Enhanced opsonophagocytic activity and inhibition of adhesion and invasion by FliC-Omp22-immune sera

Given the importance of opsonophagocytic activity in eliminating systemic A. baumannii infection, we compared the effectiveness of sera from mice vaccinated with the FliC-Omp22 or Omp22 vaccines against those injected with FliC or PBS. FliC-Omp22-immunized sera significantly improved opsonophagocytic killing of A. baumannii compared to Omp22-immunized sera (P < 0.05; Fig. 5). Specifically, at dilutions of 1:10 and 1:100, FliC-Omp22-immunized sera exhibited 59% and 45% opsonophagocytic killing activity, respectively (P < 0.05). Omp22-immunized sera also significantly enhanced opsonophagocytic killing of activity at the same dilutions, there were no significant differences between the FliC and PBS controls in this regard (P > 0.05; Fig. 5).

Fig. 5
figure 5

Comparative analysis of the opsonophagocytic killing activity and the inhibition of A. baumannii adhesion and invasion by sera from immunized and non-immunized mice. The data are represented as the mean ± SD. Datasets marked with distinct superscript letters exhibit statistical differences from each other.

Full size image

In addition to enhancing opsonophagocytic activity, FliC-Omp22-immune sera demonstrated a strong inhibitory effect on the adhesion and invasion of A. baumannii in A549 cells. At a dilution of 1:10, FliC-Omp22-immunized sera reduced both adhesion (55%) and invasion (42%) of A. baumannii compared to the controls (P < 0.05; Fig. 5). The inhibitory effect was dose-dependent, with greater reductions observed at 1:10 than at 1:100 dilutions (P < 0.05). Similarly, Omp22-immunized sera at 1:10 and 1:100 dilutions also significantly reduced adhesion and invasion compared to controls, though no significant differences were found between the FliC and PBS groups (P > 0.05).

FliC-Omp22 immunization increased survival and reduced the bacterial load of infected mice

To assess the effectiveness of the FliC-Omp22 vaccine against the lethal challenge by A. baumannii, we examined the survival rates of both immunized and non-immunized mice. FliC-Omp22-immunized mice exhibited significantly higher survival rates of 75% (6/8) than the other experimental groups and were protected from fatal A. baumannii infection (P < 0.05; Fig. 6A). In contrast, all mice in the control group succumbed to A. baumannii infection within two days, resulting in 100% mortality. Furthermore, the Omp22-immunized mice displayed a survival rate of 33% (3/8) post-infection (P < 0.05; Fig. 6A). Furthermore, we evaluated the effectiveness of the FliC-Omp22 vaccine in reducing bacterial colonization in the liver, spleen, and lungs of mice challenged with A. baumannii, and compared it with that of mice immunized with Omp22, FliC, or PBS (Fig. 6B). Notably, mice immunized with FliC-Omp22 exhibited the lowest bacterial load among those challenged with A. baumannii. Immunization with FliC-Omp22 resulted in the most significant reduction (up to 102-fold) in the liver, spleen, and lung bacterial counts following A. baumannii infection (P < 0.05; Fig. 6B), surpassing the Omp22 group. Moreover, immunization with the Omp22 antigen led to nearly 103- to 104-fold reductions in the liver, spleen, and lung bacterial loads compared to mice injected with FliC and PBS (P < 0.05). No significant differences were observed in the liver, spleen, and lung bacterial loads between the FliC and PBS controls (P > 0.05; Fig. 6B).

Fig. 6
figure 6

Active immunization with FliC-Omp22 protected mice against lethal challenges with A. baumannii. (A) On day 56, 21 days after the last immunization, the mice (n = 8 mice/group) were intraperitoneally challenged with A. baumannii and monitored twice daily for seven days. (B) Active immunization with FliC-Omp22 notably reduced bacterial burdens in the liver, spleen and lung, assessed 16 h post-infection. (C) Passive immunization with FliC-Omp22-immunized sera protected mice against lethal challenge by A. baumannii (n = 8 mice/group). (D) Passive immunization with FliC-Omp22-immunized sera resulted in a significant reduction in bacterial burdens in the spleen. Survival rates were monitored for seven days. *P < 0.05 and **P < 0.01. Datasets marked with distinct superscript letters exhibit statistical differences from each other.

Full size image

Furthermore, we assessed the efficacy of passive immunization using FliC-Omp22-immunized sera against lethal A. baumannii sepsis. As illustrated in Fig. 6C, mice injected with FliC-Omp22-immunized sera displayed the highest survival rates following challenge with A. baumannii (50% survival rates, P < 0.05), whereas all mice injected with FliC or PBS immunized sera succumbed to lethal A. baumannii infection within one day (100% mortality). Half of the mice injected with Omp22-immunized mouse sera survived after A. baumannii infection (Fig. 6C). Moreover, mice administered FliC-Omp22-immunized sera demonstrated the greatest reduction in bacterial burden compared to the other groups (P < 0.01; Fig. 6D). The bacterial load in the spleen of Omp22-immunized sera-treated mice was significantly lower than that in FliC- and PBS-immunized sera-treated mice (P < 0.05). No significant differences were observed among the FliC and PBS controls in spleen bacterial loads (P > 0.05).

Discussion

The increasing prevalence of multi- or pan-resistant strains of A. baumannii, combined with their diverse virulence factors, presents significant challenges in healthcare24. Consequently, there is an urgent need for alternative strategies such as vaccination. Omp22 is a promising antigen candidate for subunit vaccine development against A. baumannii infections due to its high sequence conservation, essential role in pathogenesis, and demonstrated protective efficacy against a broad range of clinical isolates6,13. Omp22, a 217-amino acid protein, is highly conserved across 851 A. baumannii strains, with over 95% sequence identity, as described in the previous study6. BLASTp and ConSurf analyses further confirm its conservation across diverse isolates, highlighting its potential as a universal target for vaccines or therapeutics. This conservation suggests that Omp22 could elicit protective immune responses across various strains, including those with multidrug resistance10,11. However, achieving effective immune protection against A. baumannii with the Omp22 antigen often necessitates its combination with adjuvants, conjugation to vaccine carriers, or utilization of advanced vaccine delivery platform7,14,15. This approach is particularly important because naturally occurring anti-Omp22 antibody titers in the general population may be too low to provide adequate protective benefits. Enhancing these titers through strategic vaccine formulations is key to boosting the immune response and achieving robust protection against A. baumannii infections. Although data on these antibodies are limited, it is hypothesized that exposure to non-pathogenic or commensal Acinetobacter species may induce their production in some individuals. However, the exact titers and functional capabilities of these antibodies in preventing A. baumannii infections remain poorly understood, highlighting the need for further research to assess their potential protective benefits and how they can be enhanced through vaccination. Monitoring anti-Omp22 antibody levels in at-risk populations could provide valuable translational data regarding exposure and immune status. A comprehensive understanding of the anti-Omp22 antibody landscape is essential for the development of efficacious vaccines and therapeutic interventions against A. baumannii. Furthermore, investigating the interactions between naturally occurring antibodies and enhanced vaccine formulations is crucial for optimizing immunization strategies25,26,27.

In the present study, we successfully showed that incorporation of Omp22 into the hypervariable D3 domain of flagellin significantly boosts the immune response and provides protection against lethal A. baumannii infection in a murine sepsis model. The observed improvement in immune response can be attributed to flagellin’s dual role as both a carrier and an adjuvant within the FliC-Omp22 fusion construct. This dual functionality facilitates efficient antigen display to antigen-presenting cells (APCs), thereby enhancing the overall immunogenicity. Additionally, the immunostimulatory properties of flagellin amplify the immune response by activating APCs, leading to robust adaptive immunity and enhanced protection against lethal A. baumannii infection27,28. Remarkably, the FliC-Omp22 vaccine demonstrated significant efficacy at a low dose of 10 µg, with increased survival rates among mice infected with a multidrug-resistant A. baumannii strain (58ST). This contrasts with previous Omp22 vaccines, which, although effective, required higher doses that were associated with cytotoxicity in mammalian cells7. The flagellin fusion approach addresses this limitation by leveraging TLR5 signaling, a pathway active in various cell types including epithelial cells, monocytes, and dendritic cells, to augment the immune response even at reduced antigen dosages19. Immunoassays confirmed that the FliC-Omp22 fusion protein effectively presented the Omp22 antigen while retaining flagellin’s adjuvant properties. The fusion construct exhibited the expected molecular weight and antigenic characteristics of both components, validating its structural integrity and potential as a vaccine candidate for A. baumannii infections7,29.

The Flagellin-Omp22 vaccine has demonstrated significant effectiveness in generating a Th2-biased immune response. This was evidenced by the elevated levels of IL-4 compared to IFN-γ produced by splenocytes from immunized mice upon Omp22 stimulation. The increased IL-4 levels promote the class switching of IgG isotypes to IgG1 antibodies, which are crucial for opsonization, neutralization, and protection against A. baumannii18,30. Similarly, the substitution of the hypervariable domain of flagellin with the V and F1 antigens of Yersinia pestis led to potent antigen-specific humoral immune responses, conferring protection against lethal challenges in animal models31. Similarly, substituting the hypervariable domain of flagellin with the HIV-1 p24 antigen resulted in an IgA-biased antibody response while simultaneously attenuating systemic inflammatory responses28. Incorporating the HIV gp41 membrane-proximal external region (MPER) into FliC also significantly augmented MPER-specific antibody responses in a TLR5-dependent manner29. Likewise, immunizing with four tandem copies of conserved influenza HA2 integrated into the hypervariable region of flagellin elicited robust antibody responses, providing complete protection against various lethal challenges posed by influenza A viruses in a murine model32. Additionally, replacing the hypervariable region of flagellin with a viral envelope protein from the dengue virus elicited both T cell-dependent and T cell-independent antibody responses33. Incorporating the influenza M2 protein into the flagellin molecule is safe and well-tolerated in a human clinical trial while also enhancing the production of anti-M2e-specific antibodies, indicating its potential as a universal influenza vaccine34. The FliC-Omp22 vaccine aligns with these findings, demonstrating its effectiveness by generating elevated levels of IgG1/IgG2a antibodies and increasing the IL-4/IFN-γ ratio. This supports flagellin’s known adjuvant properties in promoting Th2-type immune responses25,38,50, which are crucial for effective vaccine-induced protection against A. baumannii11,35. In our study, the reduction in IgG levels against FliC in mice immunized with the FliC-Omp22 construct likely reflects a strategic shift in the immune response. Our primary objective was to enhance the humoral response against Omp22, and the data indicate that the FliC-Omp22 construct successfully achieved this goal. The diminished IgG response to FliC may result from antigen competition, where Omp22 elicited a more robust and targeted immune response, aligning with our focus on Omp22 as a key antigen in combating A. baumannii36. Additionally, the integration of Omp22 into the construct may have altered FliC’s epitope availability, further concentrating the immune response on Omp2216,37. These findings highlight the FliC-Omp22 construct’s potential to generate a focused and effective humoral response, making Omp22 the primary target, which is essential for developing a successful multivalent vaccine.

The FliC-Omp22 vaccine not only elicits a Th2-type immune response but also initiates a Th1-type response, as evidenced by elevated IFN-γ levels in the splenocytes of vaccinated mice. This cytokine facilitates IgG class switching to IgG2a antibodies, which are associated with macrophage activation, enhanced antigen presentation, and improved bacterial clearance. These effects contribute to the establishment of durable protective immunity against A. baumannii infection. Furthermore, the enhanced proliferation rate observed in FliC-Omp22 immune splenocyte cultures upon stimulation with the Omp22 antigen indicates a robust antigen-specific memory response25. This suggests that FliC-Omp22 immune splenocytes are primed to respond rapidly to subsequent A. baumannii exposures, thereby providing long-term protective immunity38. Similar findings have been observed in other studies, such as the immunization of mice with a chimeric malaria antigen fused with the hypervariable region of truncated flagellin, which also stimulated lymphocyte proliferation and facilitated the generation of memory T-cells26. The cellular immune response elicited by the FliC-Omp22 vaccine, characterized by IFN-γ production and enhanced splenocyte proliferation, synergizes with the antibody response to improve opsonization and clearance of A. baumannii by phagocytes30,39,40.

In vitro studies of the FliC-Omp22 vaccine showed that sera from immunized mice effectively inhibit A. baumannii adhesion and invasion, and enhance opsonic killing. These findings indicate that the vaccine stimulates the production of antibodies against the Omp22 protein, which blocks bacterial attachment to host cells, facilitates opsonization, and prevents infection35,41,42. The adjuvant properties of flagellin are crucial for these effects. Flagellin promotes Th2-type immune responses, leading to antibodies that target Omp22 and impede bacterial adherence and invasion25,32,38,50,51. This results in a significant reduction in bacterial load in critical organs, highlighting the vaccine’s effectiveness in preventing systemic A. baumannii infections35,41,42. While our study effectively utilized survival rates and bacterial load measurements in key organs (liver, spleen, lungs) as established indicators of vaccine efficacy, we recognize that including additional analyses such as pathological histology and immunological markers would have provided a more comprehensive evaluation of the FliC-Omp22 chimeric vaccine’s protective efficacy. These limitations underscore the need for future research to incorporate such detailed assessments to fully elucidate the vaccine’s protective mechanisms and enhance our understanding of its impact.

Passive immunization with FliC-Omp22-immunized sera provided significant protection against lethal A. baumannii infection, as demonstrated by elevated survival rates and reduced bacterial dissemination7,43. The mechanism of protection provided by FliC-Omp22-immunized sera involves antibody-mediated opsonization and inhibition of bacterial adhesion and invasion to host cells42,44. This finding further emphasizes flagellin’s adjuvant effect in enhancing effective antibody responses targeting Omp22 when integrated into the fusion construct18,27. Active immunization with the FliC-Omp22 vaccine provides more durable protection than passive immunization. It promotes long-term antibody production and memory B cell formation, ensuring extended defense against A. baumannii25,45. Active immunization elicits both cell- and humoral-mediated immune responses, providing a comprehensive protection44,46. Conversely, passive immunization delivers only temporary protection due to the lack of cellular immunity39,47.

While the present study establishes the protective efficacy of polyclonal immune sera in passive immunization, the development of monoclonal antibodies (mAbs) targeting the highly conserved Omp22 protein across diverse A. baumannii strains may offer critical insights into the mechanisms underlying protective immunity. Despite the recognized role of Omp22 in immune responses, there is a paucity of research on mAbs specific to this protein. The development of such mAbs could significantly enhance opsonophagocytosis and provide effective protection against A. baumannii infections. Understanding how these mAbs mediate protection, particularly through opsonophagocytosis and the neutralization of bacterial virulence factors, is crucial for evaluating their therapeutic potential. A. baumannii predominantly affects a narrowly defined yet highly susceptible patient population, including those in intensive care units, immunocompromised individuals, and patients undergoing invasive medical procedures. Although the global incidence of A. baumannii infections is relatively low, the severity and increasing prevalence of multidrug-resistant strains in these high-risk groups underscore the urgent need for preventive measures. In light of the escalating threat posed by multidrug-resistant A. baumannii strains, which primarily impact these vulnerable populations, the exploration of mAbs represents a potentially immediate and targeted therapeutic strategy. This approach not only addresses critical clinical demands but also presents a more economically viable pathway for pharmaceutical development. Future investigations should prioritize the feasibility of mAb-based interventions as either complementary or alternative strategies to vaccine development in the fight against A. baumannii infections. Given the high conservation of Omp22 and its potential as a target for both vaccine and antibody-based therapies, further research into this area is imperative to combat the growing threat posed by A. baumannii.

In conclusion, integrating the Omp22 protein into the hypervariable D3 domain of flagellin holds great promise for enhancing the efficacy of an A. baumannii vaccine. FliC-Omp22 fusion protein can elicit robust Th2 immune responses against Omp22, characterized by increased IgG1/IgG2a and IL-4/IFN-γ levels. Vaccination with FliC-Omp22 also demonstrated a notable capacity to inhibit A. baumannii adhesion and invasion while enhancing opsonic killing activity. These synergistic effects lead to bacterial loads in vital organs and improved survival rates in MDR A. baumannii-infected mice, highlighting the vaccine’s promise in mitigating infections, especially amidst rising antibiotic resistance. Despite these promising results, our study highlights the need for further comprehensive assessments to fully evaluate the vaccine’s efficacy. Future research should include pathological histology and immunological marker analyses across various A. baumannii isolates. Additionally, exploring the vaccine’s applicability in other infection models, such as pneumonia and wound infections, will be crucial. The potential of the FliC-Omp22 vaccine to provide long-term underscores its significance in reducing the substantial morbidity and mortality associated with A. baumannii infections globally.

Materials and methods

Bacterial strains

Acinetobacter baumannii ATCC19606 and Salmonella enterica serovar Typhimurium PTCC1735 (S. Typhimurium) were used to produce recombinant Omp22 and flagellin (FliC) recombinant proteins, respectively. Molecular cloning and expression of the recombinant proteins were performed using Escherichia coli strains DH5α and BL21 (DE3), respectively. Escherichia coli strains, A. baumannii ATCC19606 and a previously characterized multidrug-resistant (MDR) hospital strain of A. baumannii strain 58ST35 were obtained from the Department of Biology, Shahed University (Tehran, Iran). Salmonella enterica serovar Typhimurium was kindly provided by the Bacteriology Department of the Pasteur Institute of Iran.

Ethics statement

All protocols involving animal and human cell experiments were approved by the Research Ethics Committees of Shahed University (approval number: R.SHAHED.REC.1400.197). In summary, female 6–8 weeks BALB/C mice (Pasteur Institute of Iran) were allocated to experimental and control groups, each consisting of three to five mice for serum antibody titration, in vitro lymphocyte proliferation and cytokine assays, and bacterial load assays. Survival studies of infected mice included 8 mice per group. The number of mice assigned to each group was determined using power analysis, set at 80% power, with a significance level of P < 0.05. For in vivo experiments, mice were anesthetized using a combination of ketamine and xylazine. Euthanasia was performed by exposing the mice to carbon dioxide )CO2( inhalation from a pressurized tank. Polymorphonuclear leukocytes (PMNs) were isolated from heparinized venous blood with the consent of healthy laboratory personnel and purified using the Ficoll-Hypaque density gradient method48. A549 human non-small cell lung cancer and RAW 264.7 murine macrophage cell lines were obtained from the Iranian Biological Resource Center (Tehran, Iran).

Conservation analysis of Omp22 using BLASTp and ConSurf

The conservation analysis of Omp22 in A. baumannii involved retrieving the Omp22 amino acid sequence (WP_001202415.1) from the National Center for Biotechnology Information (NCBI) Protein database. BLASTp was used to compare the sequence against a non-redundant protein database filtered for A. baumannii, focusing on globally diverse isolates. Search parameters were set for high sensitivity, using an expected threshold of 10, a word size of 3, and the BLOSUM62 scoring matrix. The sequence alignment showed that 98% of the Omp22 sequence is identical across isolates from different geographic regions, indicating a high level of conservation. The majority of the aligned sequences showed over 95% sequence identity, with complete coverage across the full length of the protein (Supplementary Fig. S3A).

Following BLASTp, ConSurf analysis was performed to evaluate the evolutionary conservation of Omp22. ConSurf provided conservation scores for each amino acid residue, based on the multiple sequence alignment and the phylogenetic relationships among the sequences. Scores were categorized from 1 (high variability) to 9 (high conservation). The evolutionary conservation scores indicated that most residues in Omp22 are highly conserved, with 75% of the residues scoring 8 or 9, corresponding to highly conserved regions (Supplementary Fig. S3B).

Bioinformatics analysis of FliC-Omp22 chimeric protein

The amino acid sequences of FliC (GenBank Accession No: AAL20871.1) and Omp22 proteins (GenBank Accession No: ABO11316.1) were retrieved from the NCBI. Following the removal of the predicted signal peptide region using SignalP 5.1 (https://services.healthtech.dtu.dk/services/SignalP-5.0/), computational tools were employed to analyze the FliC-Omp22 chimeric protein. ProtParam (http://web.expasy.org/protparam/) and IEDB tools (http://tools.iedb.org/bcell/) were used to determine the physicochemical properties of the chimeric protein, as summarized in (Supplementary Table S1). SYMPRED (https://www.ibi.vu.nl/programs/sympredwww/), PSIPRED 4.0 (http://bioinf.cs.ucl.ac.uk/psipred), and SSPro (https://download.igb.uci.edu/sspro4.html) servers were used to predict the α-helix, β-sheet, and coil structures within the FliC-Omp22 chimeric protein (Supplementary Fig. S4). BepiPred-1.0 (http://tools.iedb.org/bcell/), BepiPred-2.0 (https://services.healthtech.dtu.dk/services/BepiPred-2.0/) ABCpred (http://webs.iiitd.edu.in/raghava/abcpred/), LBtope (http://crdd.osdd.net/raghava/lbtope/), BcePred (http://crdd.osdd.net/raghava/bcepred/), and BCPred (http://ailab.cs.iastate.edu/bcpreds/) were used to predict potential linear B-cell epitopes within the chimeric protein sequence. The Robetta online server (http://robetta.bakerlab.org/) was used to generate the 3D structure of the FliC-Omp22 chimeric protein. The predicted structure was refined and validated using the QMEAN tool (https://swissmodel.expasy.org/qmean/). ElliPro (http://tools.iedb.org/ellipro/) was used to identify both linear and discontinuous epitopes on the surface of the FliC-Omp22 chimeric protein, based on its predicted 3D structure (Supplementary Fig. S5). The VaxiJen server (http://www.ddg-pharmfac.net/vaxijen/VaxiJen/VaxiJen.html/) was used to assess the antigenicity of the FliC-Omp22 chimeric protein.

Construction of Omp22, FliC, FliC-Omp22 recombinant plasmids

Specific primers (Supplementary Table S2) were designed for the fliC and omp22 genes using the genomic DNA of S. Typhimurium LT2 and A. baumannii ATCC17978 as templates, respectively. Pfu DNA polymerase-amplified omp22 and fliC amplicons were purified, digested, and ligated into the NdeI-BamHI and NcoI-HindIII (Thermo Scientific, UK) sites of the pET-28a vector (Novagen, USA), respectively. Overlap extension PCR was used to construct the FliC-Omp22 fusion fragment. Briefly, full-length fliC was amplified from the verified pET-28a/fliC vector in two fragments to remove the D3 hypervariable region for insertion of the omp22 gene. The first fragment (F1) was amplified from the first 184 amino acid residues of FliC using a forward primer with a flanking 5’ NcoI site and a reverse primer with a 5’ sequence overlapping the 5’ sequence of the omp22 gene (Supplementary Table S2). The second fragment (F2) was amplified from residues 285–494 of FliC using a forward primer with a 3’ sequence overlapping the 3’ end of the omp22 gene and a restriction site for HindIII at the 5’ end (Supplementary Table S2). Then, F1, omp22, and F2 fragments were overlapped using the conventional splicing by overlap extension (SOEing) PCR method49. Briefly, fragments were combined in equimolar amounts and subjected to PCR with 10 cycles of the program (1 min at 95 °C, 45 s at 50 °C, and 2 min at 72 °C) for overlap extension without primers, which produced the full-length FliC-Omp22 DNA, and then external primers were added to the mixture, and PCR was performed (50 °C for 1 min, 30 cycles of 95 °C for 30 s, 50 °C for 45 s, and 72 °C for 5 min). The purified chimeric DNA fragment was subsequently ligated into the NcoI-HindIII sites of the pET-28a vector. Recombinant plasmids were verified by DNA sequencing.

Preparation of recombinant proteins

Recombinant plasmids were transformed into E. coli BL21 (DE3) cells, followed by induction with IPTG (Sigma-Aldrich, USA) at 37 °C for 4 h. Subsequently, bacterial cultures were harvested, treated with 20 mM Tris-HCl (pH 8.0), sonicated on ice, and centrifuged. The resulting pellet, containing inclusion bodies, underwent resuspension in cold buffer solutions A (2 M urea, 20 mM Tris-HCl, 0.5 M NaCl, 2% Triton™ X-100; pH 8.0) and B (20 mM Tris-HCl, 0.5 M NaCl, 5 mM imidazole, 6 M guanidine hydrochloride, 1 mM 2-mercaptoethanol; pH 8.0), sonication, and centrifugation steps. The sample was then mixed with Ni-NTA agarose in buffer B, applied to a column, and washed with buffer solutions B and D (20 mM Tris-HCl, 0.5 M NaCl, 20 mM imidazole, 1 mM 2-mercaptoethanol, 6 M urea; pH 8.0). Urea removal was achieved through a stepwise decrease in concentration using buffer D. Finally, the recombinant proteins were eluted using buffer E (20 mM Tris-HCl, 0.5 M NaCl, 250 mM imidazole, 1 mM 2-mercaptoethanol; pH 8.0) and subsequently dialyzed against phosphate-buffered saline (PBS) pH 7.4 overnight. SDS-PAGE and western blotting with an anti-His tag antibody assessed the identification and correctness of the recombinant proteins48. The PYROGENT™ Plus Gel Clot LAL Assay (LONZA, USA) was used to analyze the endotoxin levels of each protein, which were found to be below 10 EU/mL.

In vitro assessment of TLR5-specific signaling

The examination of TLR5-specific signaling triggered by FliC-Omp22 was carried out through an in vitro bioassay employing two distinct cell lines: RAW 264.7 cells lacking TLR5 transfection exhibited negligible tumor necrosis factor alpha (TNF-α) synthesis upon flagellin stimulation, whereas RAW 264.7 cells transfected with TLR5 exhibited notable responsiveness to flagellin-induced TNF-α synthesis. RAW 264.7 cells were cultured in 24-well plates with RPMI 1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco) and then treated with different concentrations of either FliC-Omp22 or FliC for 4 h. Subsequently, the culture supernatant from each well was collected and subjected to TNF-α quantification using a specific ELISA assay (PeproTech, USA).

Mouse immunizations and bacterial challenges

BALB/c mice were subcutaneously administered 10 µg of Omp22, FliC, or FliC-Omp22 recombinant proteins at various sites on days 0, 14, and 28. Control mice were injected with PBS. Serum samples were obtained from orbital sinus blood to analyze antigen-specific antibodies. Infection challenge experiments were carried out using A. baumannii 58ST, as described previously7. On day 56, the mice were intraperitoneally (i.p.) challenged with a lethal dose (LD) of A. baumannii 58ST (1.9 × 107 CFU) with 10% porcine mucin (w/v; Sigma-Aldrich). Sixty hours post-infection, the spleen, lung, and liver were collected, weighed, homogenized, and cultured for bacterial load quantification. Mice were observed and monitored for seven days to assess their survival rates.

For passive immunization, BALB/c mice were intravenously administered pooled sera from immunized and non-immunized mice (100 µL) and then subjected to A. baumannii strain 58ST challenge described above. Sixty hours post-infection, spleen samples were collected and cultured on LB plates to assess the bacterial loads. Daily monitoring of survival rates continued for one week.

ELISA and Western blot assays of FliC-Omp22 fusion protein

ELISA and western blot assays were performed to confirm the identity, integrity, and binding capacity of the FliC-Omp22 fusion protein to anti-Omp22 or anti-FliC-specific antibodies, as described previously22,33. Different concentrations of FliC-Omp22 recombinant protein were coated onto 96-well ELISA plates (Nunc, USA) overnight at 4 °C in the carbonate-bicarbonate buffer. The plates were washed with PBS containing 0.5% Tween-20 (PBS-T) and blocked with PBS containing 3% bovine serum albumin (BSA, Sigma-Aldrich) for 2 h at room temperature (RT). Next, anti-Omp22- or anti-FliC-specific antibodies were added and incubated for 2 h at RT. Subsequently, the HRP-conjugated goat anti-mouse antibody (Sigma-Aldrich) was added and incubated for 1 h at RT. After the addition of TMB substrate (Invitrogen, USA) and subsequent termination of the enzymatic color reaction with 2 N H2SO4, the absorbance at 450 nm was measured.

Additionally, an immunoblotting assay was used to assess the specificity and reactivity of anti-Omp22 or anti-FliC-specific antibodies. Recombinant Omp22, FliC, and FliC-Omp22 proteins were electrophoresed and then transferred onto PVDF membranes (Amersham, USA). Subsequently, the PVDF membranes were incubated with anti-Omp22 or anti-FliC-specific antibodies, followed by HRP-conjugated goat anti-mouse antibodies. Following the addition of 3,3’-diaminobenzidine substrate (DAB, Sigma-Aldrich), the color-developed blots were recorded using a digital camera.

Evaluation of antigen-specific antibody responses

To evaluate humoral responses, we measured IgG titers against whole-cell A. baumannii, Omp22, FliC, and FliC-Omp22 antigens, as described previously48. The ELISA plates were coated with antigens, washed with PBS-T, and blocked with PBS containing 3% BSA. Serially diluted mouse sera were then added, washed with PBS-T, and incubated with HRP-conjugated goat anti-mouse IgG. The absorbance at 450 nm was measured after the addition of the TMB substrate and termination of the enzymatic color reaction. Serum IgG1 and IgG2a subtypes were determined as described above using rabbit primary anti-mouse IgG1 or IgG2a antibodies (BioLegend, USA) followed by HRP-conjugated goat secondary anti-rabbit IgG antibody (Sigma-Aldrich). The endpoint titer was determined as the highest dilution at which the optical density at OD450 exceeded that of the background wells by at least 0.1. The background wells contained PBS instead of the serum samples50.

Lymphocyte proliferation and cytokine assays

Two weeks after the last immunization, splenocytes from experimental mice were removed, homogenized, suspended in cold Hanks’ balanced salt solution (HBSS, Sigma), and subsequently treated with RBC lysis buffer (BioLegend) to remove red blood cells. Splenocytes (1 × 106 cells/mL) were resuspended in RPMI 1640 supplemented with 5% fetal bovine serum (FBS) (both from Gibco) in 96-well or 24-well plates (Nunc) and incubated with Omp22, FliC, or FliC-Omp22 antigens for 72 h. Phytohemagglutinin-A (PHA, Gibco) and unstimulated cells in RPMI 1640 were used as the positive and negative control groups, respectively. Subsequently, the culture supernatant from each well was collected and analyzed for interferon-γ (IFN-γ) and interleukin 4 (IL-4) levels using a specific ELISA assay (PeproTech). MTT (3-(4,5-dimethylthiazol-2-yl)2,5-diphenyl tetrazolium bromide) assay was performed to assess the proliferation of mouse splenocytes stimulated with antigens, as described previously51. MTT solution was added to each well and incubated at 37 °C for 4 h. Dimethyl sulfoxide (DMSO, Sigma-Aldrich) was then added to each well to dissolve the purple formazan crystals. Absorbance was measured at 550 nm, and the stimulation index (SI) was calculated by dividing the average absorbance of stimulated cells by the average absorbance of unstimulated cells (negative control).

Adhesion and invasion assays

The effectiveness of FliC-Omp22-immunized sera in inhibiting the adhesion and invasion of A. baumannii 58ST on the A549 cell line was assessed, as described previously35,52. Briefly, A. baumannii 58ST was exposed to different dilutions of heat-inactivated serum from experimental mice for 1 h at 37 °C. A549 cells were infected at a multiplicity of infection (MOI) of 100 for 2 h at 37 °C and subsequently lysed with PBS containing 0.1% Triton X-100 to evaluate bacterial adhesion. To determine bacterial invasion, infected A549 cells were treated with gentamicin (250 µg/mL) to eradicate extracellular bacteria and dissolved in PBS with 0.1% Triton X-100. Diluted samples of adhesive and invasive lysates were cultured on LB agar plates for bacterial quantification.

Opsonophagocytic assays

Acinetobacter baumannii 58ST (106 CFU) was treated with different dilutions of heat-inactivated mouse serum and infant rabbit serum for 30 min. Afterward, the suspension was added to PMNs at a concentration of 105 cells/mL and incubated for 90 min, followed by serial dilution and culture on LB agar plates for bacterial quantification41. The opsonic activity was calculated using the following formula:

Opsonophagocytic activity (%) = [1 − (bacterial CFUs of immune serum at 90 min/bacterial CFUs of pre-immune serum at 90 min)] × 100.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8 (GraphPad Software Inc., USA). Data were analyzed using a two-way analysis of variance with Tukey’s multiple comparison test. Survival data were analyzed using the Mantel-Cox log-rank test. Results are expressed as the mean ± Standard Deviation (SD), and a p-value of < 0.05 was considered statistically significant.