Induction of Broad Immunity against Invasive Salmonella Disease by a Quadrivalent Combination Salmonella MAPS Vaccine Targeting Salmonella Enterica Serovars Typhimurium, Enteritidis, Typhi, and Paratyphi A

Abstract

Bloodstream infections in low- and middle-income countries (LMICs) are most frequently attributed to invasive Salmonella disease caused by four primary serovars of Salmonella enterica: Typhi, Paratyphi A, Typhimurium, and Enteritidis. We showed previously that a bivalent vaccine targeting S. Typhi and S. Paratyphi A using a Multiple Antigen-Presenting System (MAPS) induced functional antibodies against S. Typhi and S. Paratyphi. In the current study, we describe the preclinical development of a first candidate quadrivalent combination Salmonella vaccine with the potential to cover all four leading invasive Salmonella serotypes. We showed that the quadrivalent Salmonella MAPS vaccine, containing Vi from S. Typhi, O-specific Polysaccharide (OSP) from S. Paratyphi A, S. Enteritidis and S. Typhimurium, combined with the Salmonella-specific protein SseB, elicits robust and functional antibody responses to each of the components of the vaccine. Our data indicates that the application of MAPS technology to the development of vaccines targeting invasive forms of Salmonella is practical and merits additional consideration.

1. Introduction

Invasive Salmonella disease is the most common cause of blood stream infections in low- and middle-income countries (LMICs), but it is much more than just typhoid; it encompasses enteric fever (typhoid and paratyphoid A) and invasive nontyphoidal Salmonella disease. Four key serovars of Salmonella enterica are responsible for a large majority of invasive Salmonella disease: Salmonella enterica serovars Typhi, Paratyphi A, Typhimurium, and Enteritidis. Invasive nontyphoidal Salmonella (iNTS) disease has a high associated case fatality rate and particularly affects infants and immunocompromised individuals in sub-Saharan Africa. Although conjugate vaccines against Salmonella Typhi are licensed and World Health Organization (WHO)-prequalified, there are currently no such vaccines for invasive non-typhoidal Salmonella (iNTS) disease and paratyphoid A. This study outlines the preclinical development of the first candidate quadrivalent combination Salmonella vaccine, aiming to cover all four major invasive Salmonella serotypes. The focus is on generating components against S. Typhimurium and S. Enteritidis to complement the components against S. Typhi and S. Paratyphi A described in our previous report [1].

iNTS disease is caused predominantly by serovars Salmonella enterica, serovar Typhimurium, and Salmonella enterica serovar Enteritidis [2]. Globally, over 70% of all human isolates of iNTS are S. Typhimurium and S. Enteritidis [3,4]. Poor sanitation and water treatment are frequently the cause of fecal–oral transmission of iNTS. In children under 5 years old, 12–28% of the incidence of iNTS disease is fatal [5]. There is therefore an urgent need for a vaccine targeting iNTS, particularly for pediatric populations. Thus, the development of a vaccine against both S. Typhimurium and S. Enteritidis holds the potential to substantially decrease the morbidity and mortality associated with iNTS disease.

One proposed vaccine strategy is to generate antibodies to polysaccharide components of the organisms. Salmonella Typhi has a capsular polysaccharide, Vi, which is the main component of seven licensed and WHO-prequalified conjugate vaccines against the organism [6]. While equivalent capsular polysaccharides are not present on other serovars of Salmonella enterica, there are O-specific polysaccharides (OSPs) on the surface of the latter which are components of lipopolysaccharide. An OSP of S. Paratyphi is being evaluated as a vaccine candidate [6]. S. Typhimurium and S. Enteritidis have unique O-specific polysaccharides that elicit antibody-mediated bacterial clearance [7,8,9,10,11].

In addition, genetically conserved proteins on the bacterial surface have also been proposed as potential vaccine targets, several of which are being studied preclinically. One antigen of particular interest is the serodominant secreted effector protein, SseB [12,13,14]. SseB is genetically conserved across Salmonella species and localized to the bacterial surface [13,15,16]. SseB has been used in a candidate subunit vaccine to protect chickens from infection [16,17].

In the present studies, we describe the use of a platform vaccine technology called MAPS (for Multiple Antigen-Presenting System), which, by combining both OSP and SseB, may offer important immunological advantages for protection against Salmonella. MAPS combines a rhizavidin-containing fusion protein antigen and a biotinylated polysaccharide to generate an immunogenic complex [18]. A 24-valent MAPS vaccine against Streptococcus pneumoniae has successfully completed Phase 1/2 clinical trials, attesting to the tolerability and immunogenicity of the platform [19] (NCT03803202). We have also previously demonstrated that a MAPS vaccine consisting of the Vi capsular polysaccharide for S. Typhi and the O-polysaccharide of S. Paratyphi A generates functional immune response [1]. This work led us to study MAPS complexes targeting S. Typhimurium and S. Enteritidis. We present here different MAPS complexes using a Salmonella specific protein, SseB, and OSP from S. Enteritidis J73 and S. Typhimurium S12. We evaluated the protective function of antibodies generated by MAPS and demonstrated that these Salmonella MAPS complexes generate functional antibodies in different formulations. We show that a quadrivalent Salmonella MAPS vaccine, containing OSP from S. Paratyphi A, S. Enteritidis, S. Typhimurium and S. Typhi, Vi, combined with SseB, elicits robust and functional antibody responses to each of the components of the vaccine.

2. Materials and Methods

Materials: Aluminum phosphate (alum) was obtained from Brenntag North America (Reading, PA, USA). Vi polysaccharide was provided as a gift by Dr. Szu from the NIH [20]. Adipic acid dihydrazide (ADH), 1-Ethyl-3-[3-dimethylaminopropyl] carbodiimide Hydrochloride (EDC), N-hydroxysulfosuccinimide (NHS), and 1-cyano-4-dimethylaminopyridinium tetrafluoroborate (CDAP) were purchased from Thermo Fisher (Waltham, MA, USA). Restriction endonucleases and T7 shuffle expressing competent cells were purchased from New England Biolabs (Ipswich, MA, USA). Plasmid pETDuet and other reagents were obtained from Sigma (St. Louis, MO, USA). iNTS strains were obtained from the Global Enteric Multicenter Study collection through Dr. Sharon Tennant at the University of Maryland.

Protein Purification. DNA fragments encoding His-SseB and SseA (the chaperone protein for SseB) or His-IpaB and IpgC (the chaperone protein for IpaB) were synthesized and inserted into a pETDuet vector containing Rhavi through restriction enzyme digestion and ligation. Plasmids confirmed by sequencing were introduced into E. coli T7 shuffle express cells. Transformants with the respective cloned proteins were cultivated to OD600 = 1 at 25 °C, followed by induction of protein expression with 0.2 mM IPTG at 16 °C overnight. After centrifugation, cell pellets were resuspended in lysis buffer (20 mM Tris-HCl, 500 mM NaCl, pH 8.0) and then sonicated to lyse. The supernatant was passed over a Ni-NTA column, and SseA or IpgC was washed off the column with 0.05% LDAO in a lysis buffer. Rhavi-SseB and Rhavi-IpaB proteins were then eluted in a buffer containing 250 mM Imidazole. Elutions containing the proteins were combined and subjected to purification over a gel-filtration column to isolate dimer fractions.

Purification of OSP from S. Paratyphi, S. Typhimurium, and S. Enteritidis. GMP-grade Vi was purchased from Walvax Inc. (Kunming, China). OSPs were purified from S. Paratyphi 9150 (ATCC) and clinical strains of S. Typhimurium and S. Enteritidis using a protocol established previously, with modifications [21]. Briefly, bacteria were resuspended in 6% acetic acid and boiled for 4 h at 100 °C. Supernatant was extensively dialyzed against water and then lyophilized. Resuspended OSP was treated with Q cassette and ethanol precipitation, dialyzed, and then loaded on a gel-filtration column to separate the OSPs with larger sizes. The final OSP was lyophilized to concentrate and frozen in −20 °C. Concentration was determined by the Anthrone method [22].

Biotinylation of OSP. S. Typhimurium, S. Enteritidis, and S. Paratyphi A OSPs were biotinylated with CDAP using a protocol described previously [18]. Briefly, OSP was diluted to a concentration of 2 mg/mL and incubated while stirring with CDAP (100 mg/mL) for 30 s and then with 50 mM Sodium borate for 2 min. Amine-PEG3-biotin (40 mg/mL) was added at a ratio of 1:1 (w:w) and stirred for 2 h at room temperature. Reaction was terminated by the addition of glycine to 20 mM final concentration. Biotinylated OSP underwent extensive saline dialysis before its utilization in MAPS assembly. Vi was biotinylated using Amine-PEG3-Biotin, following a previously described method, with slight modifications [18]. In summary, Vi was reconstituted to a concentration of 5 mg/mL in buffer A (0.2 M MES, 150 mM NaCl, pH 5.8). EDC (100 mg/mL in buffer A) and NHS (100 mg/mL in Dimethylformamide) were added to the solution and allowed to react for 15 min at room temperature. The solution’s pH was adjusted to pH 7.0 by adding 1 M NaHCO3 (pH 10). Amine-PEG3-Biotin (40 mg/mL in water) was introduced in a 1:1 (w:w) ratio. The reaction continued to be stirred for an additional 2 h at room temperature before the addition of glycine to achieve a final concentration of 20 mM. Biotin concentration was determined by a No-Weigh HABA/avidin Premix Biotin Quantification Kit. Vi PS was determined by acridine orange assay. Briefly, samples and standard (0.4 mg/mL) were serially diluted. Acridine orange was diluted to 1:10 dilution, added to each sample, and read at 490 nM.

MAPS formulation. All MAPS were assembled by rotating the relevant components at 4 °C overnight and purifying via size-exclusion chromatography. S. Typhimurium and S. Enteritidis MAPS were assembled at a 4:1 (w:w) protein: polysaccharide ratio and purified with size-exclusion columns. Vi MAPS were assembled at a 2:1 ratio and ParaOSP MAPS were assembled at a 6:1 ratio. Protein concentration was determined by the BCA method (Pierce), and OSP concentration was determined using the anthrone method [22]. Vi PS concentration was determined by acridine orange assay. A MAPS complex with one protein type and one polysaccharide type was considered a monovalent MAPS. Multiple monovalent MAPS could be combined into one immunization to generate immune responses against multiple Salmonella serovars and were referred to as bivalent and quadrivalent MAPS.

Antigen preparations and immunization procedures. In MAPS immunization experiments, vaccines were combined with aluminum phosphate (alum) at the specified concentration in a 5 mL tube. This mixture was tumbled overnight at 4 °C to facilitate adsorption one day prior to immunization. Rabbit immunization experiments were carried out at Cocalico Biologicals Inc. (Stevens, Hoboken, NJ, USA). All animal experiments were approved by the local Institutional Animal Care and Use Committee.

Enzyme-linked immunosorbent assay (ELISA). IgG antibody titers against S. Typhimurium OSP, S. Enteritidis OSP, S. Typhi Vi, S. Paratyphi A OSP, and the fusion proteins IpaB, SseB, and CP1 were measured using the methods described previously [23,24]. Briefly, NUNC Maxisorp plates were coated with 100 µL of 1 µg/mL (protein) or 10 µg/mL (polysaccharide) in PBS overnight at room temperature. Plates were washed with PBS Tween (PBST) and then blocked with 200 µL of PBS 1% BSA for 1 h. Serum was serially (1:3, starting from 1:50) diluted in 100 µL of PBST and incubated for 2 h at room temperature. After washing, secondary antibody (anti-rabbit IgG Fc HRP conjugate (Jackson ImmunoResearch Catalog #111-035-046)) was diluted 1:10,000 in PBST and incubated for 1 h. Plates were washed and developed with Sureblue. The reaction was stopped with 1 M HCL and read at 450 nM.

Opsonophagocytic assays (OPAs). S. Typhi OPAs was performed as has been described previously [20,24,25], with S. Typhimurium strains carrying an empty vector (Strain C5) or expressing Vi polysaccharide on the surface (Strain C5.507) [23]. The quadrivalent sera were absorbed with 25 µg/mL of S. Typhimurium OSP to eliminate the background killing of the C5.507 strain by anti-Typhimurium antibody before the assay. Activity of antisera against S. Typhimurium and S. Enteritidis was also analyzed via OPAs. Briefly, these assays were performed by first incubating target bacteria with the dilutions of heat-inactivated rabbit serum to allow the antibody to bind, after which differentiated HL60 cells and complement were incubated with the bacteria and antibodies. The reaction was terminated, and samples were plated on blood agar plates. Colonies were counted to represent the bacteria that had survived the killing. The inverse of the lowest concentration of serum in which 50% of killing occurred was defined as the killing titer.

Bactericidal assays. SBAs for S. Paratyphi were carried out as has been described previously, using the ATCC 9150 strain [26]. Bactericidal assays were set up in the same manner as the opsonophagocytic assays described above, except that differentiated HL60s were not used. The killing titer analysis is the same as for OPAs.

3. Results

3.1. Comparison of Different Fusion Proteins in S. Enteritidis MAPS Complex

Three fusion proteins were tested to assess which was most immunogenic with respect to the Salmonella OSP from S. Enteritidis. In addition to Rhavi-SseB, we also included Rhavi-IpaB (an analog of SseB in all Shigella species), as well as CP1 (Rhavi-SP1500-SP0785), which was used previously for the bivalent Typhi and Paratyphi A vaccine [1]. MAPS constructs were formulated at a 4:1 protein-to-polysaccharide ratio. Each immunization contained 5 µg polysaccharide/dose. In Figure 1A, the three MAPS vaccines were administered to rabbits in two doses spaced three weeks apart. Blood sera were collected at three different time points: before the initial injection (pre, solid symbols), three weeks after the first injection (P1, open symbols), and three weeks after the second injection (P2, half-open symbols). ELISAs against the two antigens of the vaccine (OSP and relevant protein) were performed to analyze serum IgG antibody titers.

Rabbits that received any of the three MAPS constructs had approximately a 1000-fold increase in antibody titer to the OSP compared to baseline after two doses (Figure 1A). Rabbits that received SseB-S. Enteritidis MAPS vaccines had the highest average antibody titers, statistically significantly higher than titers from those immunized with MAPS vaccines, including Rhavi-IpaB (p = 0.032) or CP1 (p = 0.016). There was no statistically significant difference in the response to vaccines containing Rhavi-IpaB and CP1. All three fusion proteins were immunogenic, as expected (Figure 1B).

Next, we evaluated functional antibody responses. Opsonophagocytic assays (OPAs), instead of bactericidal assays, were used to assess the functionality of MAPS-induced antibodies with S. Enteritidis OSP due to its reproducibility across different isolates (Figure 1C) [27]. Immunized rabbit sera were assessed post one (P1) and post two (P2) immunizations for their ability to opsonize and enable phagocytic killing of S. Enteritidis bacterial cells (Figure 1C). Rhavi-SseB-S. Enteritidis MAPS generated a similar killing titer after two immunizations (Figure 1C, squares) compared to S. Enteritidis MAPS associated with Rhavi-IpaB or CP1. Rhavi-SseB-S. Enteritidis MAPS-immunized sera demonstrated a similar increase in killing activity between the first and second doses, which suggests that two doses are necessary for this vaccine to be most effective. Of note, we observed positive correlations between the ELISA titer and killing titers of the antisera for S. Typhimurium (Spearman test r = 0.9524, p = 0.0011) and S. Enteritidis (Spearman test r = 0.8333, p = 0.015) MAPS after two immunizations (Figure S1).

We conclude that, of the three fusion proteins evaluated, Rhavi-SseB resulted in the highest immune response induced by Salmonella MAPS vaccines. Based on this, we conducted all subsequent studies using Rhavi-SseB as the fusion protein.

3.2. Role of Fusion Protein Rhavi-SseB

A potential benefit of using a Salmonella-specific protein as the fusion protein is the ability to provide additional (polysaccharide-independent) protection against infection. To evaluate this possibility, we studied the ability of protein-directed antibodies to kill the organism. Rabbits were immunized twice with 100 µg protein of Rhavi-SseB on a two-week schedule. OPAs were performed with the post two immunizations sera and S. Typhimurium and S. Enteritidis as the target bacteria (Figure 2A, open symbols). We observed killing activity by Rhavi-SseB sera compared to pre-immunization sera (Figure 2A, closed symbols) for both bacteria, demonstrating that antibodies to Rhavi-SseB could provide some additional protective activity against the two Salmonella serovars.

We then tested the function of protein-directed antibodies generated following immunization with MAPS vaccines, using sera from animals immunized with MAPS vaccines made with fusion protein CP1 for comparison. As shown previously, rabbits immunized with SseB-containing MAPS vaccines generated higher antibody concentrations to the polysaccharides than did those with CP1 as fusion protein. To adjust for this, we compared the killing activity in OPA per arbitrary unit of OSP antibody (Figure 2B). For S. Typhimurium, the results demonstrated that the killing titer per unit of OSP antibody was significantly higher in sera from Rhavi-SseB MAPS-immunized rabbits compared to those immunized with CP1 MAPS (p = 0.008). However, such a difference was not noted when killing activity against S. Enteritidis was analyzed in the same fashion. These data suggest that a pathogen-specific fusion protein may enhance killing activity of sera from immunized animals for certain organisms.

3.3. Analysis of S. Typhimurium MAPS

Monovalent MAPS constructs were generated using an OSP from S. Typhimurium and Rhavi-SseB as the fusion protein. First, we assessed the production of antibodies against the OSP in rabbits immunized with Rhavi-SseB-S. Typhimurium MAPS construct after the first and second immunizations. After two immunizations, the rabbits generated a robust antibody response to S. Typhimurium OSP (Figure 3A). The antibody response to the fusion protein Rhavi-SseB was also measured after two immunizations, and similar antibody titers were individually generated by the Rhavi-SseB-S. Typhimurium MAPS construct compared to the Rhavi-SseB-S. Enteritidis MAPS construct in Figure 1 (Figure 3B). The functionality of the OSP antibodies present after two doses of MAPS vaccines was then tested with opsonophagocytic assays. S. Typhimurium MAPS vaccines generated antibodies with robust killing activity (Figure 3C). Together, the data from Figure 1 and Figure 3 show how both MAPS constructions produced potent antibodies against the vaccine’s antigens and showed the ability to eradicate bacteria carrying the same OSP.

3.4. Multivalent MAPS against Two and Four Salmonella Species

To target several clinically relevant Salmonella species in a single vaccine, multivalent MAPS vaccines were constructed by incorporating monovalent MAPS. We tested a bivalent MAPS construct (containing S. Typhimurium and S. Enteritidis OSP) and a quadrivalent MAPS (containing S. Typhimurium, S. Enteritidis, S. Paratyphi A OSP, and S. Typhi Vi) both using Rhavi-SseB as the fusion protein.

We tested the antibody production in rabbits immunized with the bivalent MAPS vaccine. Robust antibody responses were generated against both OSPs, at similar levels to those achieved by each monovalent MAPS vaccine (Figure 4A compared to Figure 1A and Figure 3A). These results suggest little to no interference in antibody production with the addition of another MAPS construct in the vaccine. We then assessed OSP antibody production by the quadrivalent MAPS vaccine formulation. After two doses of the quadrivalent MAPS vaccine, robust antibody titers were detected for all four polysaccharides (Figure 4B). Both multivalent MAPS vaccines showed high levels of antibody production which were similar to those for Rhavi-SseB after two doses (Figure 4C).

The ability of vaccine-generated antibodies to target and kill the four Salmonella species was tested with opsonophagocytic assays for all organisms except S. Paratyphi A, for which we used a bactericidal assay. For the bivalent MAPS vaccine, killing activity against S. Typhimurium and S. Enteritidis was very similar to the killing titers generated with each of the monovalent MAPS vaccines (Figure 5A compared to Figure 1C and Figure 3C). The quadrivalent MAPS vaccine containing all four polysaccharides was tested for the functionality of all the antibodies generated (Figure 5B). The killing titers that were generated against all four Salmonella species were similar to the monovalent and bivalent MAPS vaccines’ counterparts (Figure 5B compared to Figure 1C and Figure 3C) [1]. These data suggest that a quadrivalent MAPS vaccine can generate robust and functional antibodies against all four Salmonella strains.

4. Discussion

Salmonella is a type of facultative intracellular pathogen demonstrating the capability to survive both extracellularly and within monocytes and macrophages [28]. These attributes present a challenge in preventing Salmonella infections, as the bacteria exhibit a remarkable ability to evade the immune system.

Several vaccine techniques to target iNTS have been tested in preclinical trials (reviewed in [6]). A live-attenuated S. Typhimurium strain, WT05, was tested in a Phase 1 clinical trial, in which it elicited an immune response and was well-tolerated in healthy volunteers [29,30]. A potential concern, however, was that the bacteria were shed in the stool for up to 23 days after immunization, which poses a risk of infection in vulnerable people and an increased potential for transfer of genetic material [29,30]. S. Typhimurium have also been genetically modified to generate anti-LPS antibodies in mice, with limited fecal shedding [30,31,32,33].

Another approach is the use of outer membrane vesicles (OMVs), as, for example, the Generalized Modules for Membrane Antigens (GMMA) platform being used by the GSK Vaccines Institute for Global Health (GVGH) [9,34]. GMMA employs immunogenic, scalable OMV-based components that have been tested in Phase I and Phase II clinical trials for Shigella sonnei (NCT02676895 and NCT03527173) [30,35,36,37,38] and are now being evaluated for iNTS in Phase I and II studies (NCT05480800).

Several studies have demonstrated that antibodies against S. Typhimurium and S. Enteritidis OSP are protective in mice [7,9,10,11,30,38,39,40,41,42,43]. Studies of sera collected from healthy individuals identified antibodies to the OSP of S. Typhimurium that had bactericidal activity [44,45,46]. Taken together, these data suggest that eliciting B-cell responses to these antigens may represent a viable vaccine strategy. Experimental O-antigen-based conjugate vaccines of S. Typhimurium and S. Enteritidis protected mice from lethal Salmonella infections [9,41,43,47,48,49]. The Center for Vaccine Development and Global Health has developed a conjugate vaccine for S. Typhimurium and S. Enteritidis. A vaccine based on core and O-specific polysaccharides (OSPs) chemically linked to flagellin protein FliC was immunogenic and protective in mice and has completed a Phase 1 clinical trial (NCT03981952) [8,30,48]. As a parallel approach, surface-exposed, genetically conserved Salmonella proteins have been studied for their potential to generate protection from infection. Members of the outer membrane protein family (OmpCDF), siderophores, and proteins from the type III secretion system are immunogenic in mice [5,50,51,52].

In this work, we designed two MAPS constructs to target S. Typhimurium and S. Enteritidis using O-specific polysaccharides from both organisms. We determined that the Salmonella-specific protein SseB was an excellent fusion protein in these constructs (Figure 1), resulting in high antibody titers to OSP and excellent killing activity. We focused on SseB due to its conserved nature, its location on the cell surface, and its role in infection [5,14,53,54] (Figure 3, Figure 4 and Figure 5). We also observed that SseB can potentially provide additional protection for a strain of S. Typhimurium (Figure 2). Our approach has unique features compared to others. In addition to the robust antibody responses to OSP, MAPS also generated functional antibodies against the important Salmonella antigen SseB. However, the role of anti-SseB antibodies in the protection against Salmonella serovar still needs to be confirmed in future clinical trials.

We designed bivalent and quadrivalent Salmonella MAPS vaccines and demonstrated that combining MAPS into one immunization can potentially provide protection against multiple organisms. The bivalent MAPS vaccine containing OSPs for S. Typhimurium and S. Enteritidis demonstrated significant antibody production and killing activity, comparable to their monovalent counterparts (Figure 4 and Figure 5). The quadrivalent MAPS vaccine against S. Typhimurium, S. Enteritidis, S. Typhi Vi, and S. Paratyphi A induced production of significant functional antibodies (Figure 4 and Figure 5). The data demonstrate that protection could be potentially provided for all four serovars of Salmonella enterica with two doses of this vaccine. This would eliminate the need for multiple differing immunizations in endemic areas where vaccine compliance may be difficult [6]. Overall, our data indicate that applying MAPS technology to vaccine development against invasive forms of Salmonella is feasible and warrants further consideration and confirmation of protection by challenge studies.

5. Conclusions

Our study showed that a combinational MAPS vaccine containing Vi, OSPs from S. Typhimurium, S. Enteritidis, and S. Paratyphi A, and conserved Salmonella protein SseB generated functional antibodies against all the components included in the vaccine. The results support further development of the vaccine.