Towards a Four-Component GMMA-Based Vaccine against Shigella

Shigellosis remains a major public health problem around the world; it is one of the leading causes of diarrhoeal disease in low- and middle-income countries, particularly in young children. The increasing reports of Shigella cases associated with anti-microbial resistance are an additional element of concern. Currently, there are no licensed vaccines widely available against Shigella, but several vaccine candidates are in development. It has been demonstrated that the incidence of disease decreases following a prior Shigella infection and that serum and mucosal antibody responses are predominantly directed against the serotype-specific Shigella O-antigen portion of lipopolysaccharide membrane molecules. Many Shigella vaccine candidates are indeed O-antigen-based. Here we present the journey towards the development of a potential low-cost four-component Shigella vaccine, eliciting broad protection against the most prevalent Shigella serotypes, that makes use of the GMMA (Generalized Modules for Membrane Antigens) technology, a novel platform based on bacterial outer membranes for delivery of the O-antigen to the immune system.


Introduction
Shigella spp. are a leading cause of diarrhoeal mortality among all ages worldwide [1][2][3]. The spectrum of disease ranges from mild diarrhoea to severe dysentery (bloody, mucoid diarrhoea) with fever. Recent estimates report approximately 270 million diarrhoea episodes due to Shigella, with around 212,000 deaths among all ages per year, of which 64,000 are in children under 5 years. Among these deaths, 90% occur in low-and-middle-income countries (LMICs) [1]. Although mortality rates from diarrhoeal diseases have decreased since 1990, diarrhoea morbidity remains high, particularly in LMICs, where access to health care, relevant microbiological diagnostics, quality of water, and sanitation are poor, and there is poor access to adequate health-care facilities, diagnostics, and therapeutic interventions. Complications from infection are especially common in malnourished children, with possible long-term consequences including stunting and cognitive impairment in children experiencing multiple diarrhoeal episodes [4]. Additionally, diarrhoea due to Shigella can cause significant morbidity among travellers and military recruits. Antibiotic resistance of Shigella is increasing, and conventional therapeutic antibiotics against shigellosis have become progressively less efficient. Many outbreaks of shigellosis have been reported due to resistant strains [5,6], and the WHO's Global Antimicrobial Resistance Surveillance System has identified Shigella as a priority pathogen for the development of new interventions [7].

Generalized Modules for Membrane Antigens Platform
Gram-negative bacteria (e.g., Shigella) have an outer membrane and an inner membrane linked through multiple protein systems. They naturally shed small amounts of outer membrane vesicles (OMV) by a process of blebbing. Indeed, spontaneous release of vesicles from bacteria results in the formation of native OMV, but for many species this occurs at very low levels, insufficient to be purified for inclusion in a vaccine formulation. To overcome low yields, Gram-negative bacteria can be genetically manipulated to greatly increase the shedding of outer membranes [31].
A reverse vaccinology study on N. meningitidis identified genes yielding a hyperblebbing phenotype and first suggested a possible application of OMV as a platform for vaccine development [32,33]. While the genes identified in the N. meningitidis reverse vaccinology study were specific to N. meningitidis, a forward genetic approach based on genome-wide transposon screening in E. coli laboratory strain identified multiple null mutations that generate a similar hyper-blebbing phenotype. This included disruption of the Tol-Pal system, which is highly conserved across Gram-negative bacteria [34]. In a subsequent study, a tolR null mutation was induced in extraintestinal E. coli (ExPEC) pathogenic strains and first demonstrated immunogenicity and protection conferred by OMV generated with this technology [35]. Following studies expanded this approach as a generic platform for homologous and heterologous protein expression system for vaccine applications [36].
Shigella sonnei was engineered to overexpress OMV via tolR null mutation, and additional mutations were added to minimize the capacity of GMMA to promote a reactogenic response when delivered in a parenteral formulation [31]. This approach allowed reactogenicity to be efficiently reduced, while maintaining the immunopotentiator effect of the Toll-like receptor 4 (TLR4) triggered by lipid A [37][38][39]. Additional mutations were developed to remove antigens detrimental to a vaccine or to enhance safety (e.g., ∆virG) [40].
GMMA can be produced at high yields using a simple and robust process of manufacturing, potentially leading to low-cost vaccines [51]. In fact, after fermentation of the mutated bacteria, GMMA are purified with a rapid filtration-based downstream processing. GMMA can be produced at high yields using simple, robust, and cost-effective manufacturing processes and were quickly recognized as an ideal technology for low resource settings, where the ability to produce and deliver vaccines sustainably in markets without strong commercial drivers is critically important and where poverty-related conditions such as enteric and diarrhoeal diseases are among the leading causes of mortality in children.
Importantly, GMMA mimic the outer membrane of bacteria, where key antigenic components to elicit an immune response are lacking the ability to cause associated disease. They contain pathogen-associated molecular patterns (e.g., lipopolysaccharides, lipoproteins) conferring self-adjuvanticity and have an optimal size for immune stimulation, with potential for strong immunogenicity [30].
On the other hand, GMMA are quite complex, and a comprehensive panel of analytical methods has been put in place to characterize GMMA as particles and for analysis of key antigens displayed on their surface to ensure quality and consistency of manufacture at release and to follow GMMA stability over time [52][53][54].

Preclinical Experience
The first S. sonnei strain was genetically modified to increase blebbing (∆tolR), produce penta-acylated lipid A (∆htrB) with reduced endotoxicity, and stably express the virulence plasmid encoding for the immunodominant OAg component of the LPS (by replacing the plasmid-encoded virG gene with E. coli nadA and nadB genes to remove the nicotinic acid auxotrophy) [40]. The scale-up of the process, consisting of two simple tangential flow filtration steps after fermentation of the mutated bacteria [31], to GMP pilot scale yielded a very robust process that generates approximately 100 mg of GMMA proteins per litre of fermentation, meeting predetermined product specifications.
As GMMA contain LPS and lipoproteins, assessing potential reactogenicity was prioritized for vaccine development. In an in vitro monocyte activation test (MAT), S. sonnei GMMA showed a markedly lower IL-6 production compared with peripheral blood mononuclear cells (PBMC) stimulated with the parent GMMA with unmodified lipid A [40]. It was found that the residual activity of GMMA was largely due to non-lipid A-related TLR2 activation [39]. Additionally, a modified rabbit pyrogenicity test was established, based on the European Pharmacopoeia pyrogens method, but using intramuscular administration of the full human dose (100 µg of protein) of S. sonnei GMMA adsorbed on Alhydrogel (1790-GAHB) [40].
Aluminium hydroxide was used as adsorbent to further reduce potential reactogenicity of GMMA particles [55]. Minor temperature increases were observed in rabbits receiving the vaccine, but these were transient and within an acceptable range. A repeat dose toxicology study in rabbits using intramuscular (IM, 6.4/100 µg/dose), intranasal (IN, 5.1/80 µg/dose), or intradermal (ID, 0.64/10 µg/dose) administration routes showed good tolerability of the vaccine by all routes, supporting its suitability for use in humans [40]. As for the rabbit toxicology studies, the Irwin test in rats showed no evidence that IN immunization with 1790GAHB could cause neurological findings. In mice, 1790-GMMA formulated on Alhydrogel elicited anti-LPS specific serum IgG in a dose-dependent manner. Vaccination of mice with 1.75 ng of OAg/29 ng protein gave an already significant anti-LPS IgG response after a single injection, with a strong increase following a second injection 4 weeks later [40]. At the doses tested in the rabbit toxicology study, the antibody response peaked after the first dose IM, while after the second dose IN and ID.
Strategies applied to S. sonnei strain for GMMA production were demonstrated also for S. flexneri strains [39,56]. S. flexneri serotypes, in contrast with S. sonnei and with the exception of S. flexneri 6, share a common OAg backbone. To assess the potential for immunologic cross-reactivity of S. flexneri GMMA, 14 different S. flexneri GMMA producer strains (∆tolR) were generated [56]. Corresponding GMMA were tested in mice for their ability to induce cross-reactive and cross-functional antibodies. GMMA, eliciting antibodies cross-binding (by FACS) and cross-killing (by serum bactericidal assay, SBA) on numerous heterologous S. flexneri serotypes were identified, suggesting that broadly protective Shigella vaccines are possible and may cover multiple serotypes with a limited number of components. For example, S. flexneri 1b GMMA elicited antibodies able to bind and kill different heterologous serotypes, including S. flexneri serotypes 4a and 6. On the other hand, narrow specificity GMMA immunogens were identified, such as S. flexneri 2a, inducing antibodies able to bind to and kill only the homologous serotype bacteria or few others. These data informed the design of a four-component vaccine based on GMMA from S. sonnei, S. flexneri 1b, S. flexneri 2a, and S. flexneri 3a ( Figure 1). The resulting formulation tested in mice and rabbits elicited strong SBA titres against a large panel of Shigella serotypes [56]. exception of S. flexneri 6, share a common OAg backbone. To assess the potential for immunologic cross-reactivity of S. flexneri GMMA, 14 different S. flexneri GMMA producer strains (ΔtolR) were generated [56]. Corresponding GMMA were tested in mice for their ability to induce cross-reactive and cross-functional antibodies. GMMA, eliciting antibodies cross-binding (by FACS) and cross-killing (by serum bactericidal assay, SBA) on numerous heterologous S. flexneri serotypes were identified, suggesting that broadly protective Shigella vaccines are possible and may cover multiple serotypes with a limited number of components. For example, S. flexneri 1b GMMA elicited antibodies able to bind and kill different heterologous serotypes, including S. flexneri serotypes 4a and 6. On the other hand, narrow specificity GMMA immunogens were identified, such as S. flexneri 2a, inducing antibodies able to bind to and kill only the homologous serotype bacteria or few others. These data informed the design of a four-component vaccine based on GMMA from S. sonnei, S. flexneri 1b, S. flexneri 2a, and S. flexneri 3a ( Figure 1). The resulting formulation tested in mice and rabbits elicited strong SBA titres against a large panel of Shigella serotypes [56].
This study also highlighted the importance of anti-OAg antibodies for binding and inducing bactericidal activity against OAg-positive bacteria. Actually, antibodies generated against OAg-negative GMMA were unable to bind to OAg-positive bacteria. These results agreed with the recent work of Mancini et al. [42], demonstrating that anti-protein antibodies are induced in mice upon immunization with either OAg-negative or OAgpositive S. sonnei GMMA, but the presence of OAg chains on the bacterial surface shield the bacteria from anti-protein antibody binding, and therefore anti-OAg antibodies are the main drivers of bactericidal activity against OAg-positive bacteria. This finding is reinforced by the functional analysis of human sera from a phase 2b study of the S. sonnei GMMA-based vaccine [57]. The adsorption of anti-OAg antibodies from post-vaccination sera confirmed that anti-protein antibodies were not able to induce complement-mediated bactericidal killing against S. sonnei OAg-positive bacteria.  This study also highlighted the importance of anti-OAg antibodies for binding and inducing bactericidal activity against OAg-positive bacteria. Actually, antibodies generated against OAg-negative GMMA were unable to bind to OAg-positive bacteria. These results agreed with the recent work of Mancini et al. [42], demonstrating that anti-protein antibodies are induced in mice upon immunization with either OAg-negative or OAg-positive S. sonnei GMMA, but the presence of OAg chains on the bacterial surface shield the bacteria from anti-protein antibody binding, and therefore anti-OAg antibodies are the main drivers of bactericidal activity against OAg-positive bacteria. This finding is reinforced by the functional analysis of human sera from a phase 2b study of the S. sonnei GMMA-based vaccine [57]. The adsorption of anti-OAg antibodies from post-vaccination sera confirmed that anti-protein antibodies were not able to induce complement-mediated bactericidal killing against S. sonnei OAg-positive bacteria.
Interestingly, antibodies that are not targeting the Oag are functional against Oagnegative bacteria, and immunodominant protein antigens were identified by proteomic analysis [42]. Little is known about OAg length and density regulation during infection and, therefore, protein exposure. It is possible that protein antigens could contribute to protection through other mechanisms [58]. Ongoing activities are focused on understanding the role of proteins in protection and whether the presence of protein antigens on Shigella GMMA represent an added value for GMMA vaccines compared with other OAgbased formulations. It has already been demonstrated that GMMA-associated proteins stimulate cooperation of B-cells with T-helper cells for induction of a T-cell dependent anti-OAg specific response. In recently published works, GMMA were shown to induce a stronger response in wild-type than in T-cell-deficient mice, clearly highlighting a strong T-dependent component of the immune response induced by GMMA [43,44]. Interestingly, when presented on GMMA, even short OAg chains are able to elicit part of their response through direct activation of B-cells without involving T-cells. This could be explained by the particulate nature of GMMA displaying highly organized and repetitive antigens (i.e., OAg) to the immune system. Such multiplicity can favour efficient B-cell receptor (BCR) crosslinking, delivering strong activation signals to the B-cells independent of polysaccharide length [59]. Moreover, GMMA contain different TLR agonists, including lipid A and lipoproteins, able to provide synergistic TLR signalling to B-cells [37], as confirmed by MAT. This could favour a rapid and strong response after one only injection [60]. Even if a purer T-dependent response could have a positive impact on memory and persistency of the antibody response, when compared head to head in a mouse study, no differences were observed in the ability of GMMA or traditional glycoconjugates to induce strong responses after re-injection or in the persistence of the antibody response induced [43]. In this study, GMMA induced higher anti-O-antigen IgG than glycoconjugate administered without Alhydrogel, as already verified with Salmonella GMMA [49]. Additional studies are currently ongoing to compare four-component GMMA and glycoconjugate formulations in animal models.
A series of studies was performed to better assess the criticality of Shigella GMMA attributes. The expertise acquired on bacteria manipulation allowed the efficient generation of strains characterized by different OAg length or different O-acetylation pattern. Shigella OAg biosynthesis is dependent on the Wzx/Wzy pathway, and the degree of repeating unit polymerization (i.e., OAg size) is regulated by the Wzz family of proteins, responsible for unique polysaccharide modal lengths [61]. OAg repeating units can also be assembled into high molecular weight capsules (Group 4 Capsules, G4C) if an additional G4C operon is present, as it is in the case of S. sonnei and S. flexneri 6 serotypes [43,44,62]. Consequently, OAg populations at different length are displayed on GMMA. Mutations were introduced in the OAg locus or in the G4C operon in order to display homogeneous polysaccharide populations of different sizes on GMMA surface. For S. sonnei, S. flexneri 2a, and S. flexneri 6, characterized by OAg with different structural characteristics (zwitterionic, neutral, and negatively charged sugars), no major impact of OAg chain length on the immune response elicited in mice was verified [43,44].
GMMA were also engineered to express OAg with different O-acetylation patterns [45]. The role that O-acetylation can have on the immune response elicited by S. flexneri OAg can be important not only to assess the protective role against homologous strains, but also against heterologous serotypes sharing common epitopes due to such modifications. S. flexneri 2a and 1b OAg with different O-acetylation patterns were displayed on GMMA by knocking-out the responsible O-acetyltransferase genes. The data obtained suggested that the OAg O-acetylation is not critical for the generation of functional antibodies against homologous or heterologous strains.

Clinical Experience with S. sonnei Monocomponent GMMA Vaccine
S. sonnei GMMA formulated on Alhydrogel (1790GAHB) has been tested in five clinical trials to date (Table 1). Two parallel phase 1 trials in European adult population (18-45 years) were performed [63]. During the two phase 1 trials, a range of doses from 1.5/25 µg (OAg/protein) to 6/100 µg were tested and different routes using IM, ID, and IN routes. The maximum dose of OAg was limited by the amount of protein that could be administered without compromising on participant safety.
The initial phase 1 study (H03_01TP, clinicaltrial.gov NCT02017899) evaluated the safety and immunogenicity profile of three IM injections at 4 weekly intervals compared with placebo. S. sonnei 1790-GMMA containing 12 µg/mL of OAg and 200 µg/mL protein (6/100 µg dose) was adsorbed to 0.7 mg Al 3+ /mL of aluminum hydroxide. Dilutions of the full dose with placebo were performed to obtain the other doses tested. Intramuscular doses ranged from 0.06/1 µg to 6/100 µg (OAg/protein content) in an injected volume of 0.5 mL. In the second phase 1 study (H03_02TP, clinicaltrial.gov NCT02034500), the safety and immunogenicity of 1790GAHB was also evaluated when delivered three times, 4 weeks apart at different doses intradermally or intranasally, testing other routes of immunization which have different theoretical advantages, in addition to a cohort including IM administration. Intradermal doses were from 0.006/0.1 µg to 0.6/10 µg in 0.05 mL and IN doses were 0.3/5 µg to 4.8/80 µg in a total volume of 0.4 mL administered as a 0.2 mL aliquot into each nostril.
In both studies, vaccines were well-tolerated after administration by IM, ID, and IN vaccination routes. Solicited events were mild to moderate in severity. Vaccine-related unsolicited adverse events were uncommon. No subjects experienced serious adverse events (SAEs). Among unsolicited adverse events, eight cases of neutropenia were reported in vaccinees [64]. All cases of neutropenia were transient and had a benign clinical outcome. In the phase 1 and subsequent trials, no pyrogenic response was observed post-vaccination. In line with the outcome from preclinical rabbit studies [40], in the phase 1 common solicited local adverse event was limited and primarily confined to injection-site pain. Rhinorrhoea was the most common adverse event by the IN route. A correlation between dose and local reactions was observed, but even at the highest dose of 6/100 µg, the vaccine had an acceptable reactogenicity profile.
When delivered by the IM route, by pooling together the responses to the 1.5/25, 3/50, and 6/100 µg doses (defined as high-dose groups), reverse cumulative antibody distribution 28 days after the second vaccination (day 85) for the high-dose group showed a similar distribution to the antibody responses measured in an adult population following natural infection, with a median of 305 ELISA units compared with 121 ELISA units for the convalescent patients and a substantial increase compared with the pre-vaccination distribution [27,63]. On the last day of the study (day 225), the median antibody in vaccinees was 241 ELISA units, higher than the median in the convalescent patients. Moreover, antibodies with complement-mediated bactericidal activity were detected in vaccinees but not in placebo recipients, and correlation between SBA titres and anti-S. sonnei LPS serum IgG levels was observed [65].
Despite screening prior to enrolment, about half of the subjects had detectable anti-S. sonnei LPS antibodies prior to vaccination. Individual antibody responses suggested qualitative and quantitative differences between subjects with and without preexisting detectable antibody. Subjects with preexisting antibody tended to have stronger responses to the vaccine and a different pattern of boosting and decay kinetics.
In contrast with preclinical animal results [40], via the ID or IN routes, the vaccine was poorly immunogenic at the doses tested. The finding of no significant response following intranasal administration is consistent with findings following the administration of another Shigella vaccine [66]. It may be that the doses administered in our trial were too low to elicit an immune response. Additional studies could be useful for exploring higher ID and IN doses in an attempt to identify the optimal immune response and to expand the use of GMMA to mucosal vaccination.
Subsequently, an extension of the initial IM phase 1 study evaluated memory response induced by booster vaccination with 1.5/25 µg of 1790GAHB in individuals who had undetectable antibodies at baseline in H03_01TP. Prior to vaccination, specific antibody levels in some subjects immunized 2-3 years earlier were well-maintained, suggesting good immune persistence from vaccination. 1790GAHB was confirmed to be well-tolerated and to induce a significant booster response in previously primed adults, even those receiving primary doses of OAg/protein content as low as 0.059/1 µg, both in terms of anti-LPS IgG response and bactericidal activity, regardless of the priming dose. Additionally, the proportion of subjects achieving convalescent antibody levels (121 EU) 7 days post booster was 71% compared with 24% in the naïve group, suggesting a good immunological memory [65,67].
Based on the results from the two phase 1 trials, the IM route was selected for further development in a phase 2a trial (H03_04TP, clinicaltrial.gov NCT02676895) in African adults aged 18-45 years. Participants received two vaccinations with the 1790GAHB vaccine at doses of either 1.5/25 µg of OAg/protein or 6/100 µg 1 month apart, or vaccination with active comparator vaccines (a quadricomponent meningococcal vaccine at day 1 and tetanus, diphtheria, and acellular pertussis vaccine at day 29). The most frequently reported solicited local and systemic adverse events across all groups were pain and headache, respectively, with a similar reporting frequency between participants receiving the 1790GAHB and active comparators. One case of severe systemic reaction was reported (severe headache after first vaccination in group 6/100 µg). Neutropenia was designated as an adverse event of special interest (AESI) to allow systematic data collection for those events, based on the cases observed in the previous studies. Ten episodes of neutropenia were reported in two (9%), three (12%), and one (4%) participants in the 1.5/25 µg, 6/100 µg, and active comparator groups, respectively. These episodes were all considered probably or possibly related to vaccination; two of them occurred in the 1.5/25 µg group (one mild and one moderate), five in the 6/100 µg group (four mild and one moderate), and three in the control group (one mild, one moderate, and one severe). All cases of neutropenia were Vaccines 2022, 10, 328 8 of 14 transient (recovering within 7 days and by study end) and asymptomatic, as confirmed by daily home visits during the 7 days post vaccination. No SAEs were reported [68].
Anti-S. sonnei LPS serum IgG geometric mean concentrations were evaluated at days 1, 29, and 57 and compared with anti-S. sonnei LPS antibody levels in convalescent patients naturally exposed to S. sonnei. Very high baseline anti-S. sonnei LPS serum IgG levels were observed in the study population; the median antibody levels were 30-fold higher than observed in Israeli convalescent sera and in healthy European adults vaccinated with the same doses of 1790GAHB. Despite this, the 1790GAHB vaccine induced robust antibody responses in Kenyan adults. At day 29, geometric mean antibody levels increased 2.10-fold and 4.43-fold from baseline in volunteers immunized with 1.5/25 µg and 6/100 µg of 1790GAHB, respectively. Day 57 antibody levels in the 1790GAHB-immunized groups were not statistically different from those at day 29. No increase was observed among controls. In the 1.5/25 µg dose group, seroresponse was 68% after first and 90% after second immunization; a 96% seroresponse was observed regardless of immunization in the volunteers receiving 6/100 µg.
The 1790GAHB studies described above provided evidence that the vaccine was well tolerated and immunogenic in both European and African adults ( Figure 2) and in the presence of preexisting antibody, vaccine-or infection-induced. the control group (one mild, one moderate, and one severe). All cases of neutropenia were transient (recovering within 7 days and by study end) and asymptomatic, as confirmed by daily home visits during the 7 days post vaccination. No SAEs were reported [68]. Anti-S. sonnei LPS serum IgG geometric mean concentrations were evaluated at days 1, 29, and 57 and compared with anti-S. sonnei LPS antibody levels in convalescent patients naturally exposed to S. sonnei. Very high baseline anti-S. sonnei LPS serum IgG levels were observed in the study population; the median antibody levels were 30-fold higher than observed in Israeli convalescent sera and in healthy European adults vaccinated with the same doses of 1790GAHB. Despite this, the 1790GAHB vaccine induced robust antibody responses in Kenyan adults. At day 29, geometric mean antibody levels increased 2.10-fold and 4.43-fold from baseline in volunteers immunized with 1.5/25 μg and 6/100 μg of 1790GAHB, respectively. Day 57 antibody levels in the 1790GAHB-immunized groups were not statistically different from those at day 29. No increase was observed among controls. In the 1.5/25 μg dose group, seroresponse was 68% after first and 90% after second immunization; a 96% seroresponse was observed regardless of immunization in the volunteers receiving 6/100 μg.
The 1790GAHB studies described above provided evidence that the vaccine was well tolerated and immunogenic in both European and African adults ( Figure 2) and in the presence of preexisting antibody, vaccine-or infection-induced. To evaluate the efficacy of the vaccine in a challenge human infection model (CHIM) for shigellosis, a phase 2b trial was performed. Additional objectives were to evaluate safety and immunogenicity of 1790GAHB in healthy US adults aged 18-45 (H03_03TP, clinicaltrial.gov NCT02676895). Two IM injections with 1.5/25 µg (OAg/protein) of 1790GAHB, Vaccines 2022, 10, 328 9 of 14 administered one month apart, or a placebo control were followed by an oral challenge of volunteers with live S. sonnei 53G, 4 weeks after the second vaccination. Although safe and immunogenic, under the trial conditions, the vaccine did not show any efficacy against shigellosis [57]. Regarding neutropenia, following the evaluation of the signal of asymptomatic neutropenia among mostly participants of African descent in the phase 1 and phase 2a studies, and the phase 2b study, which showed that the S. sonnei 1790GAHB was not more likely to cause neutropenia than other licensed vaccines, it was thus no longer considered as an AESI for further Shigella GMMA vaccine development.
Evidence was shown from the challenge trial; however, the evidence was in agreement with data previously published [26], showing that anti-LPS IgG antibody plays a significant role in protection against shigellosis (i.e., anti-S. sonnei LPS serum IgG concentration was higher among participants who did not develop shigellosis compared with participants who developed the disease following the challenge dose) (Figure 3). However, insufficient immunogenicity was considered to be the main root cause for the lack of efficacy. This finding suggested the need to enhance S. sonnei immunogenicity by increasing the OAg dose as well as by increasing the likely effect of the second vaccination, which in this trial showed no further increase in antibody levels from the levels after the first vaccination. This can be done by increasing the interval between the first and second vaccinations. To evaluate the efficacy of the vaccine in a challenge human infection model (CHIM) for shigellosis, a phase 2b trial was performed. Additional objectives were to evaluate safety and immunogenicity of 1790GAHB in healthy US adults aged 18-45 (H03_03TP, clinicaltrial.gov NCT02676895). Two IM injections with 1.5/25 μg (OAg/protein) of 1790GAHB, administered one month apart, or a placebo control were followed by an oral challenge of volunteers with live S. sonnei 53G, 4 weeks after the second vaccination. Although safe and immunogenic, under the trial conditions, the vaccine did not show any efficacy against shigellosis [57]. Regarding neutropenia, following the evaluation of the signal of asymptomatic neutropenia among mostly participants of African descent in the phase 1 and phase 2a studies, and the phase 2b study, which showed that the S. sonnei 1790GAHB was not more likely to cause neutropenia than other licensed vaccines, it was thus no longer considered as an AESI for further Shigella GMMA vaccine development.
Evidence was shown from the challenge trial; however, the evidence was in agreement with data previously published [26], showing that anti-LPS IgG antibody plays a significant role in protection against shigellosis (i.e., anti-S. sonnei LPS serum IgG concentration was higher among participants who did not develop shigellosis compared with participants who developed the disease following the challenge dose) (Figure 3). However, insufficient immunogenicity was considered to be the main root cause for the lack of efficacy. This finding suggested the need to enhance S. sonnei immunogenicity by increasing the OAg dose as well as by increasing the likely effect of the second vaccination, which in this trial showed no further increase in antibody levels from the levels after the first vaccination. This can be done by increasing the interval between the first and second vaccinations.

Moving toward a Four-Component Shigella GMMA-Based Vaccine
An alternative S. sonnei construct and S. flexneri GMMA have been developed to have higher OAg density on GMMA particle compared with the original S. sonnei GMMA, for inclusion in a four-component Shigella formulation. This will allow an increase in the dose of each OAg with no impact on safety, maintaining the total dose of proteins and lipid A below the highest dose tested in clinical trials with acceptable safety in 1790GAHB. The appropriateness of increasing the OAg dosage, with respect to the 1.5 µg OAg tested in the challenge trial with 1790GAHB, is also supported by observations carried out in previous clinical trials performed with the 1790GAHB vaccine, which showed an immunogenicity dose response [65,68], and by the much higher antigen content used for other OAg-based Shigella vaccines. The Flexyn2a vaccine, produced by Limmatech biotechnology in vivo in E. coli and containing 10 µg of Shigella flexneri 2a OAg was tested in a phase 2b proof of concept CHIM trial [69]. The Flexyn2a vaccine administered twice, one month apart intramuscularly, was found to be safe and immunogenic and demonstrated a protective efficacy point estimate of 72.4% against severe diarrhoea post-challenge. Another OAgbased candidate vaccine, the SF2a-TT15, containing 10 µg oligosaccharide conjugated to tetanus toxoid was tested in healthy adults in Israel. The vaccine was also shown to be more immunogenic at the 10 µg dose compared with the 2 µg dose [70]. Finally, the US NIH S. sonnei conjugate vaccine demonstrating 74% protective efficacy against shigellosis in adults [20] contained 25 µg of OAg conjugated to Pseudomonas aeruginosa recombinant exoprotein A (rEPA).

Conclusions
According to recent estimates, at least 2.5 billion people in LMICs lack access to improved sanitation, and over 884 million lack access to improved drinking water [71]. Global access to improved sanitation and meticulous hand washing represent the ideal solution for preventing transmission of Shigella. However, this can be difficult to achieve in all settings, and it can be challenging to sustain, given the financial and logistical constraints in low-resource regions. Vaccination represents the best solution in the short term. However, Shigella epidemiology is not static and has been shown to be time-and geography-dependent, complicating vaccine design, already challenged by a limited up to date understanding of serotype burden, need for cross-reactivity of immune responses to different serotypes to simplify vaccine composition, and possible interference among multiple vaccine components. GMMA are highly immunogenic bacterial outer membrane exosomes used as an antigen delivery system [29,30]. Combinations of GMMA, manufactured at high yield and at low cost, seem to be a promising technology for addressing the goal of an affordable vaccine against Shigella for LMICs.
Generic methods for cost effective production of GMMA and their formulation have been developed. Starting from a S. sonnei monocomponent vaccine, bacteria engineering strategies, manufacturing processes, and analytics have been easily extended to additional serotypes, with the aim of developing an OAg-based multicomponent vaccine with broad protection against the most prevalent serotypes. Furthermore, GMMA have been used as a useful tool for rapidly modifying key features of Shigella OAg, investigating the criticality of certain parameters (e.g., OAg of different size or differently decorated with O-acetyl groups) [43][44][45]. GMMA can be directly tested in animal models, allowing questions of scientific interest to be rapidly addressed.
Different OAg-based vaccines are currently in development [69,70,72]. A previous experimental parenteral S. sonnei OAg conjugate vaccine showed 71% efficacy against homologous S. sonnei infection in children 3 years of age and older but not in the younger children [21]. GMMA represent an innovative and different OAg delivery system. GMMA co-deliver several types of pathogen-associated molecular patterns that confer to GMMA self-adjuvanticity. Further than OAg, GMMA contain Shigella proteins that can contribute to efficacy through different mechanisms. Shigella strains could be further mutated to express on GMMA additional Shigella proteins (e.g., Ipa proteins, virG, absent in current preparations) to increase the breadth and magnitude of the immune response induced, resulting in increased protection against different serotypes. Additionally, heterologous antigens could be presented on GMMA, supporting the development of multicomponent vaccines, covering more pathogens at the same time (e.g., ETEC-Shigella) [73].
The Shigella GMMA 1790GAHB vaccine was shown in clinical trials to have an acceptable safety profile and to be able to induce a robust immune response in adult humans when administered IM. This demonstrates its feasibility as a vaccine against shigellosis, which at a higher OAg dose is likely to induce sufficient antibodies to protect against moderate to severe diarrhoea due to Shigella in areas of the highest need.
A four-valent GMMA-based candidate has just entered a phase 1-2 clinical trial to evaluate its safety and immunogenicity (clinicaltrial.gov NCT05073003). The vaccine will be first administered in European adults and then tested in an age de-escalation trial (from least vulnerable adult population to most vulnerable paediatric population) in a shigellosisendemic population in Africa. The results of this study will allow the selection of the most appropriate dose for further vaccine development in infants 9 months of age, which is the main target age group for this vaccine.