Impact of O-Acetylation on S. flexneri 1b and 2a O-Antigen Immunogenicity in Mice

Shigellosis is a diarrheal disease caused prevalently by Shigella flexneri and S. sonnei and representing a major global health risk, particularly in developing countries. Bacterial O-antigen (OAg) is the primary target of the host immune response and modifications of its oligosaccharide units, including O-acetylation, are responsible for the variability among the circulating S. flexneri serotypes. No vaccines are widely available against shigellosis and the understanding of the immunogenicity induced by the OAg is fundamental for the design of a vaccine that could cover the most prevalent Shigella serotypes. To understand whether a different O-acetylation pattern could influence the immune response elicited by S. flexneri OAg, we employed as a vaccine technology GMMA purified from S. flexneri 2a and 1b strains that were easily engineered to obtain differently O-acetylated OAg. Resulting GMMA were tested in mice, demonstrating not only no major impact of O-acetyl decorations on the immune response elicited by the two OAg against the homologous strains, but also that the O-acetylation of the Rhamnose III residue (O-factor 9), shared among serotypes 1b, 2a and 6, does not induce cross-reactive antibodies against these serotypes. This work contributes to the optimization of vaccine design against Shigella, providing indication about the ability of shared epitopes to elicit broad protection against S. flexneri serotypes and supporting the identification of critical quality attributes of OAg-based vaccines.


Introduction
Shigella infections are among the top causes of moderate to severe diarrhea throughout the world, causing approximately 112 million cases of which a great percentage is represented by children under 5 years of age [1].
There are four different Shigella species: S. boydii, S. dysenteriae, S. flexneri, and S. sonnei, divided in more than 50 serotypes based on antigenic variation of their O-Antigen (OAg). S. flexneri is predominant in low-and middle-income countries (LMIC) causing 66% of cases of shigellosis, with serotypes 1b, 2a, 3a, and 6 representing the most abundant ones [1][2][3]. The increasing emergence of multidrug-resistant clones makes Shigella even more worrisome, pressing the race for the development of a vaccine, still not widely available on the market [4][5][6][7]. However, several groups worldwide are making great efforts in testing candidate vaccines, most of them targeting the OAg portion of Shigella lipopolysaccharide (LPS) anchored through the lipid A to the outer membrane. LPS is the most relevant virulence factor in Shigella pathogenesis and plays an important role in bacterial resistance to both innate and adaptive immunity [8][9][10][11][12].
All S. flexneri serotypes, except S. flexneri 6, share a conserved polysaccharide backbone (also known as serotype Y) [13][14][15] consisting of the following repeating unit: 2)-α-L-RhapIII-(1,2)-α-L-RhapII-(1,3)-α-L-RhapI-(1,3)-β-D-GlcpNAc-(1-. The diversity of the S. flexneri serotypes is due to the modification of the common OAg backbone with glucosyl and/or O-acetyl (OAc) groups as a result of bacteriophages infec-tion and acquisition of OAg-modifying enzymes [16]. In the case of S. flexneri 2a OAg, the repeating unit is substituted with a glucose on RhaI, added by the glucosyltransferase GtrII (epitope shared with S. flexneri 2b) [17]. Moreover, two different O-acetyltransferases can modify the repeating unit by adding O-acetyl groups on GlcNAc (OacD) and/or on RhaIII (OacB) [18,19]. 6-O-acetylation on GlcNAc confers the antigenic group O-factor 10 [19]. S. flexneri 1b OAg repeating unit is substituted with a glucose on GlcNAc, added by the glucosyltransferase GtrI (epitope shared with S. flexneri 1a). The O-acetyltransferase OacA O-acetylates RhaI (epitope that allows discrimination with S. flexneri 1a), whereas the Oacetyltransferase OacB adds a 3/4 O-acetylation on RhaIII ( Figure 1) [13,18]. The addition of an O-acetyl group on position 3 or 4 of RhaIII provides the antigenic group O-factor 9. Wang and colleagues studied the distribution of the 3/4-O-acetylation in S. flexneri and the antigenicity that resulted from this modification. PCR screening of the oacB gene in 730 clinical isolates of S. flexneri demonstrated that the oacB-mediated 3/4-O-acetylation is widespread in serotypes 1a, 1b, 2a, and Y. Among them, 1a, 1b, and 2a are the predominant serotypes in China and other Asian countries, and the high 3/4-O-acetylation rates (>94%) found in these serotypes suggest that the presence of O-factor 9 may play a role in their preferable dissemination. However further studies are necessary to elucidate if this OAg modification contributes to the virulence of S. flexneri.
O-factor 9 has also been reported in S. flexneri serotype 6, despite its genetic determinant was found in the oacC gene (72% homologous to oacB) [20].
There are no direct evidence in the current literature about the role of O-acetylation in the immune response induced by S. flexneri OAg. It has been reported that a S. flexneri 2a OAg conjugate vaccine induced protection against S. flexneri serotype 6 in humans and the O-factor 9 was suggested as the structural basis for the observed cross-reactivity between S. flexneri 2a and S. flexneri 6 LPS [21,22].
In this study we aim to verify the criticality of O-acetylation pattern for S. flexneri OAg, in particular for S. flexneri 2a and 1b, two of the most prevalent S. flexneri serotypes, and the impact that OAc epitopes can have in the cross-recognition of homologous and heterologous S. flexneri serotypes. We used the generalized modules for membrane antigens (GMMA) technology to rapidly generate different OAg structures with the desired features. GMMA are outer membrane vesicles naturally released from genetically engineered Gramnegative bacteria, having the great advantage to display the OAg in its natural outer membrane context. Shigella bacteria are mutated in order to increase vesicles formation, through the deletion of the tolR gene [23]. Potential reactogenicity is reduced by modifying the lipid A structure usually by deletion of the htrB or msbB genes [24]. For these advantages and their easy and cost-effective manufacturability, GMMA have been proposed as a strategy for the development of a multi-component vaccine against Shigella [25][26][27][28][29]. Starting from overblebbing ∆tolR S. flexneri 1b and 2a strains, we generated knock-out mutants of S. flexneri 2a for the O-acetyltransferases genes oacB and oacD and of S. flexneri 1b for the O-acetyltransferase gene oacB. GMMA obtained from these strains were used for immunogenicity studies in mice. Notably, OAc on RhaI of S. flexneri 1b was not altered as its removal would convert the strain to serotype 1a.
The results from this study will contribute to the design of cross-protective OAg-based vaccines against Shigella flexneri.

Bacterial Strains and Generation of Mutants
S. flexneri 2a and S. flexneri 1b were acquired from Public Health England (PHE) and engineered to obtain the different mutants. The null mutations were obtained by replacing the genes of interest with an antibiotic resistance cassette by homologous recombination using the lambda red recombineering system [30]. The list of all bacterial strains generated and primers used is reported in Supplementary Table S1. The O-acetyltransferases null mutations were inserted both in a wild type and ∆tolR background to generate bacterial strains to be used for serum bactericidal assay (SBA) and GMMA purification, respectively. Absence of O-acetyltransferase genes was confirmed by PCR and the phenotype of each mutant was verified by H 1 NMR analysis of OAg extracted from the resulting GMMA, as previously described [31,32].

GMMA Production and Characterization
S. flexneri 2a and 1b GMMA were produced and purified as previously described [31]. GMMA total protein content was estimated by bicinchoninic acid assay (BCA) using BSA as a reference. The total OAg amount and sugar composition were determined by high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD), after performing acid hydrolysis directly on GMMA. In particular, the OAg amount was quantified based on the detection of rhamnose, as previously described [33,34]. The OAg extracted was characterized by high performance liquid chromatography-size exclusion chromatography (HPLC-SEC) (TSK gel 3000 PWXL column with TSK gel PWXL guard column equilibrated in 0.1 NaCl, 0.1 NaH 2 PO 4 , 5% CH 3 CN, Tosoh Bioscience, Tokyo, Japan) with differential refractive index (dRI) detection, using dextrans as the standards to estimate the molecular size distribution as previously described [31,33]. OAg structures were confirmed by 1 H-NMR analysis measured with a Bruker AvanceIII 400 spectrometer at 400 MHz and 323 K.

Mouse Studies
GMMA obtained from S. flexneri 2a and 1b, differing for O-acetylation pattern, were tested in mice. Female, five-week-old wild-type CD1 mice were immunized intraperitoneally (IP) with 200 µL of vaccine containing 0.005, 0.05, or 0.5 µg of OAg at day 0 and 28. Single sera were collected at day 27 and 42. All animal studies were performed at GSK ARC Animal Care Facility under the animal project 526/2020-PR 26/05/2020 approved by the Italian Ministry of Health. The studies were ethically reviewed and carried out in accordance with European Directive 2010/63/EEC, the GSK policy on the Care, Welfare and Treatment of Animals, and local animal welfare guidelines under Italian authorization. All sera were analyzed by ELISA for anti-S. flexneri OAg total IgG using wild type S. flexneri 1b and 2a OAg as plate-coating antigens (at 0.5 and 5 µg/mL concentration in carbonate buffer pH 9, respectively). All coating antigens were purified in house as previously described [33]. Single sera collected at day 42 were also assayed in SBA based on luminescent readout (L-SBA) [35] against S. flexneri 1a, 1b, 2a, 2b, or 6 wild-type or mutant strains, as previously described [28,31].

Statistical Analysis
Analysis was performed using GraphPad Prism 7 (La Jolla, CA, USA). The Kruskal-Wallis analysis with post-hoc Dunn's test was used to compare multiple groups. Doseresponse relationships were evaluated through Spearman's rank correlation. The parallelism of dose-response curves was assessed by the parallel line method: when the slopes of the curves obtained by log-transforming ELISA or SBA results vs. log transformed antigen doses were not significantly different from each other, comparison of the Y-intercepts was performed. GMMA were purified from each S. flexneri strain and analytically characterized (Table 1). Importantly, mutations introduced did not impact the OAg amount in S. flexneri 2a GMMA as all GMMA were characterized by similar OAg to protein ratios, estimated through HPAEC-PAD and micro BCA analyses. GMMA displaying a mutated OAg for S. flexneri 1b presented instead a slightly decreased OAg/protein ratio with respect to its wild-type counterpart.
The OAg was extracted from each GMMA and analyzed through HPLC-SEC to determine the OAg molecular weight (MW) distribution. Importantly, introduced mutations did not impact OAg size. Indeed, the four strains of S. flexneri 2a showed very similar percentage of three different OAg populations: around 15% for the high MW (HMW), around 50% for the medium MW (MMW), and around 35% for the low MW (LMW) OAg (this peak corresponds to core only and core plus few OAg repeats). Moreover, the two S. flexneri 1b strains resulted to have very similar OAg populations: 65% and 35% for the MMW and LMW, respectively.  (Figures 1 and S1). Importantly, deletion of OAc groups in specific positions did not impact the O-acetylation degree in the remaining ones, allowing a good comparison among the different constructs in mice.

Immunogenicity Study in Mice with OAg Differing for the OAc Pattern
All GMMA were compared in mice at three different OAg doses (0.5, 0.05, and 0.005 µg) and mice were immunized intraperitoneally twice at 4-week intervals.
Mice sera obtained after first (day 27) and second immunization (day 42) were analyzed through ELISA using Shigella wt fully O-acetylated OAg as coating antigen for each S. flexneri serotype. For both S. flexneri 2a and 1b, the anti-OAg IgG response induced in mice was not impacted by the OAg O-acetylation pattern, as the anti-OAg IgG obtained by immunizing with the differently O-acetylated GMMA was not statistically different neither at day 27 nor at day 42 ( Figure 2). These data were analyzed by applying the parallel line approach, which revealed no significant differences between wild type and mutated GMMA in the range of OAg doses investigated. Sera collected at day 42 were further analyzed for antibodies functionality by L-SBA performed against each of the corresponding differently O-acetylated Shigella strain. L-SBA data confirmed the results obtained by ELISA and proved that the functional activity of sera increased against each GMMA was comparable regardless the presence of OAc groups (Figure 3). The different GMMA groups were compared to each other by applying the parallel line approach, which revealed no significant differences in the slopes and Y-intercepts as shown in Figure 3. To understand weather O-acetylation can impact the ability of S. flexneri OAg to raise cross-functional antibodies against heterologous serotypes, sera obtained by immunizing mice with the highest S. flexneri 2a and 1b OAg dose were tested in L-SBA also against S. flexneri 1b and 2a, respectively as well as against other heterologous S. flexneri serotypes.
In particular, for what concerns S. flexneri 2a, besides S. flexneri 1b, we selected S. flexneri serotypes 1a and 6 that also are O-acetylated on RhaIII and serotype 2b that shares with S. flexneri 2a the glucosylation on RhaI (epitope conferring the type 2 serotype).
SBA results, reported in Figure 4, clearly show that all sera raised against S. flexneri 2a GMMA resulted in high bactericidal titers against S. flexneri 1a and 2b (IC 50 10 3 -10 5 ), while relatively low titers were measured against S. flexneri 1b and 6 (IC 50 < 10 3 ). However, the presence/absence of OAc on S. flexneri 2a OAg did not make any difference in the functionality of the antibodies raised, indicating that common OAg OAc did not play a role in inducing cross-reactive antibodies across the heterologous S. flexneri strains tested. Similarly, polyclonal sera obtained with S. flexneri 1b GMMA were tested in SBA against S. flexneri 1a, 2a, and 6 that also present OAc on RhaIII. Serum bactericidal killing of S. flexneri 1a was high (IC 50~1 0 5 ), whereas low IC 50 titers were measured when sera were tested against S. flexneri 6 and almost no killing was observed in the case of S. flexneri 2a. These results were independent from the presence of O-acetylation on S. flexneri 1b OAg.

Discussion
Shigella is a major cause of morbidity worldwide and antimicrobial resistance makes the development of a widely accessible vaccine an even higher global health priority [4]. The major challenge for the development of a vaccine against Shigella is related to the high number of serotypes causing the disease. In response to this need, multi-component vaccines are currently under development to ensure broad protection against Shigella [36,37].
It is known that structural characteristics of saccharide antigens, including O-acetylation, can play a major role on the immune response elicited and can be critical for the design of highly effective vaccines [38]. This is the case of Salmonella Typhi and Salmonella Paratyphi A, where the Vi and OAg immunogenicity were related to their degree of O-acetylation [39,40]. On the other hand, polysaccharide O-acetylation was not found to be essential for antigenicity and immunogenicity of some Streptococcus pneumoniae strains [41], however it is an aspect not to underestimate in the context of vaccine development.
Few preclinical data have been collected so far on the role that O-acetylation can have on the immune response elicited by S. flexneri OAg. This 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. Indeed, cross-reactivity could help to reduce the complexity of a Shigella vaccine composition.
GMMA represent a powerful technology to facilitate this kind of investigation, besides being a promising vaccine approach [42]. Indeed, GMMA-producing strains can be easily mutated to display OAg with the desired features and resulting GMMA can be rapidly tested in animal studies. Here we have generated GMMA displaying S. flexneri 2a and 1b OAg with different O-acetylation patterns by knocking-out the responsible genes, similarly to what has already been reported to obtain different OAg size [43].
The impact that OAg O-acetylation has on several aspects of the Shigella pathogenesis and immunogenicity has never been clarified so far. In this study we looked at the role of S. flexneri 2a and 1b OAg O-acetylation in driving an immune response in mice, when delivered with GMMA. In particular, the aim was to verify the criticality of O-acetylation occurring in two positions of S. flexneri 2a OAg, on RhaIII and on GlcNAc, and on RhaIII of S. flexneri 1b in generating a broad and functional anti-OAg immune response.
The data obtained suggested that the OAc on RhaIII or/and GlcNAc of S. flexneri 2a OAg and the OAc on RhaIII of S. flexneri 1b OAg are not critical for the generation of functional antibodies against homologous or heterologous strains.
These results are in line with a recent publication in which we showed that S. flexneri 6 OAg O-acetylation has little effect on OAg conformation and hence may not be essential for the antigenicity of serotype 6. This was corroborated by a similar in vivo study in mice, using S. flexneri 6 GMMA as O-antigen delivery systems, that shows that O-acetylation does not have an impact on the immune response elicited by the serotype 6 O-antigen [32]. Moreover, binding studies of synthetic S. flexneri 2a OAg to five protective mIgGs showed that none of the acetates affect significantly the antibody recognition [44]. However, it is fair to highlight that conclusions drawn on studies done using the mouse model may not reflect the effects of OAg O-acetylation in immune response occurring in humans. Indeed, as previously mentioned, a S. flexneri 2a conjugate vaccine was able to elicit in humans cross-reactive antibodies against S. flexneri 6 and is reasonable to attribute this result to the O-factor 9 (OAc of RhaIII), the only epitope that these two OAg structures have in common, although no cross-reactivity was observed from studies done in the mouse model, as also confirmed by our data, independently from the O-acetylation pattern [21]. Moreover, Hlozek et al. recently performed computational investigation of the effect of Oacetylation of RhaIII in serogroup 2 on OAg conformation, showing that this modification limits the flexibility of the OAg backbone, changing the dominant conformation from a C-curve to a helix [45]. This change was suggested to provide a rationale for the crossreactivity observed between antibody raised by an O-acetylated 2a vaccine and serotype 6. Finally, it is worth to remind that in the context of cross-reactivity among different Shigella serotypes, besides O-acetylation, also OAg glucosylation may play a role in generating conformational epitopes shared among different OAg structures and this aspect deserves further investigation.
Not necessarily data in mice will be predictive of data in humans, but the systematic study performed here allowed to collect a certain data package that will be corroborated from data from clinical trials soon. Different Shigella OAg-based vaccines are being tested in clinical trials and analysis of the sera for their bactericidal activity against a panel of different serotypes will allow to confirm or disprove preclinical data and help also clarifying the role of O-acetylation in the immune response elicited by Shigella OAg.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10.3390/ microorganisms9112360/s1, Figure S1: Complete 1 H NMR spectra of OAg structures of the differently O-acetylated S. flexneri strains., Table S1:List of bacterial strains and primers used in this study.