A Novel Shigella O-Polysaccharide–IpaB Conjugate Vaccine Elicits Robust Antibody Responses and Confers Protection against Multiple Shigella Serotypes

ABSTRACT Shigella is responsible for high burdens of diarrhea and dysentery globally. Children living in areas of endemicity are the most affected, and currently, there are no licensed vaccines to prevent shigellosis. Vaccine approaches have traditionally targeted the bacterial lipopolysaccharide as a protective antigen. Shigella O-polysaccharide (OPS) conjugated to recombinant Pseudomonas aeruginosa exotoxin A (rEPA) or tetanus toxoid (TT) is advanced in clinical evaluation. Adequate efficacy of these vaccines, particularly in the infant target group, remains to be demonstrated. A major limitation of the OPS-glycoconjugate concept is its limited coverage, since immunity to the O antigen is serotype specific, and there are multiple disease-causing serotypes. Another concern is the use of protein carriers already included in multiple other childhood vaccines. This study reports a novel Shigella OPS conjugate vaccine that uses the Shigella invasion plasmid antigen B (IpaB) as the carrier protein. IpaB is a virulence factor component of the Shigella type III secretion system and highly conserved among Shigella serotypes. It is robustly immunogenic and a protective antigen. IpaB and IpaB containing nonnative amino acids (nnAA) were produced at large scale using cell-free protein synthesis. Incorporation of nnAA enabled site-specific conjugation of IpaB to Shigella flexneri 2a OPS using click chemistry, yielding OPS-IpaB glycoconjugate. Parenteral immunization of mice with the OPS-IpaB vaccine resulted in high levels of OPS- and IpaB-specific serum IgG and robust protection against lethal S. flexneri 2a or Shigella sonnei challenge. The OPS-IpaB vaccine is a promising new vaccine candidate with the capacity to confer broad protection against clinically relevant Shigella serotypes. IMPORTANCE Diarrhea caused by Shigella species results in long-term disability and mortality globally, disproportionally affecting younger children living in poor countries. Although it is treatable by antibiotics, the rapid and widespread emergence of resistant strains and the highly contagious nature of the disease compel the development of preventive tools. Currently, several Shigella OPS conjugate vaccines are being evaluated in clinical studies, but these rely exclusively on immunity against the bacterial O antigen, which limits their coverage to only the immunizing serotype; multivalent vaccines are needed to protect against the most prevalent serotypes. This is the first report of a novel Shigella OPS-conjugate vaccine that uses Shigella IpaB as a carrier and protective antigen. This vaccine, administered parenterally, elicited robust immunity and protected mice against lethal infection by S. flexneri 2a or S. sonnei. The OPS-IpaB vaccine is a promising candidate for evaluation in vulnerable populations.

IpaB was produced using Vaxcyte's proprietary Xpress 1 CFPS platform (22). The native protein sequence was used as a template to generate multiple IpaB single-site variants that efficiently incorporated nnAA p-azidophenylalanine (pAMF) (Fig. 1), at specific sites purposely selected to be distant from immunologically relevant epitopes (23,24). Expression yields of both native and nnAA-containing IpaB were .200 mg/L as measured through [ 14 C]leucine incorporation ( Fig. 2A). Six of the single-site IpaB pAMF incorporation variants (highlighted in orange in Fig. 2A) were selected based on expression levels, and their pAMF sites were combined to generate 3 and 4 pAMFcontaining IpaB (variants 1 to 6) as indicated in Fig. 2B. Expression of IpaB by the 3 or 4 pAMF-containing variants compared to native IpaB was estimated using [ 14 C]leucine FIG 1 Schematic representation of the expression of nnAA-containing IpaB using XpressCF 1 CFPS technology. Cell-free protein synthesis (CFPS) for expression of nnAA (pAMF) containing IpaB was performed by generating a reaction mixture comprising (1) Escherichia coli cell extract, (2) components necessary for transcription (T7 polymerase and the DNA template), and (3) molecular machinery facilitating site-selective pAMF incorporation (synthetase, orthogonal tRNA [otRNA], and pAMF) along with (4) essential components necessary for energy production (nucleoside triphosphates [NTPs], amino acids, salts, and cofactors). Postexpression, CFPS product was harvested, and target protein (IpaB-var) was purified and analyzed by SDS-PAGE analysis. pAMF incorporation was confirmed through labeling with DBCO-TAMRA dye. var, variant. Shigella OPS-IpaB Conjugate Vaccine mSphere incorporation followed by autoradiogram analysis and was found to be largely unaffected by pAMF incorporation (Fig. 2C). Importantly, the immune reactivity of IpaB variants 1 to 4 was compared to that of native IpaB using human Shigella convalescent-phase sera and was found to be almost identical (Fig. 2D), confirming the preservation of immune reactive epitopes. IpaB-variant 1 (IpaB-var1), which produced larger conjugates in initial SEC-MALS analysis (data not shown), was selected as the carrier protein, scaled up using CFPS (performed at 2-to 4-L reaction volume in a bioreactor under controlled conditions), and produced at high purity and homogeneity. Biophysical characterization of the purified protein using SEC-MALS estimated a molar mass of 420 6 0.4 kDa (Fig. 2E), which is a close approximation to the mass of a hexamer of IpaB in solution (theoretical molecular mass of monomeric IpaB-variant 1, ;65 kDa). SDS-PAGE analysis of the purified protein revealed .95% purity, and the dibenzocyclooctyne-tetramethylrhodamine (DBCO-TAMRA) labeling confirmed the presence of pAMF sites (Fig. 2E, inset). We next performed DBCO derivatization of S. flexneri 2a OPS (OPS-DBCO) followed by click-chemistry reaction with IpaB-variant 1 to produce the OPS-IpaB conjugate vaccine (Fig. 3A). SEC-MALS analysis of OPS-IpaB revealed an average molecular mass of 285 6 2 kDa (Fig. 3B), and SDS-PAGE analysis confirmed almost complete utilization of IpaB in the conjugation reaction ( Fig. 3B, inset). In addition, OPS-DBCO was also conjugated to CFPS-produced eCRM, an nnAAcontaining variant of diphtheria toxin variant CRM 197 (25), to generate OPS-CRM for use as a control in mouse experiments. Immunogenicity of S. flexneri 2a OPS-IpaB conjugate vaccine. BALB/c mice were immunized intramuscularly (i.m.) on two occasions (days 0 and 28) with 10 mg/dose of OPS-IpaB admixed with an alum-based adjuvant, AdjuPhos. Groups receiving OPS-CRM, IpaB (both at 10 mg/dose), AdjuPhos, or phosphate-buffered saline (PBS) were included as controls. Mice immunized intranasally (i.n.) with sublethal doses of S. flexneri 2a or S. sonnei were included as positive controls for challenge experiments. OPS-IpaB and OPS-CRM conjugates elicited high levels of OPS-specific serum IgG ( Fig. 4A and B); titers were not significantly different between the two groups (Fig. 4B). As expected, high levels of OPS-specific serum IgG were detected in mice that received S. flexneri 2a i.n. (Fig. 4A). In all vaccinated groups, IgG titers increased with each vaccination and reached a plateau ;28 days after the boost ( Fig. 4A and B). Mice immunized with S. flexneri 2a OPS-IpaB also produced high levels of IpaB-specific IgG, and titers surpassed those elicited by IpaB alone (Fig. 4A and B): mean values of 2.1 Â 10 6 and 1.4 Â 10 6 enzyme-linked immunosorbent assay (ELISA) units (EU)/mL, respectively, on day 55 postvaccination. High levels of CRM-specific serum IgG were also produced by OPS-CRM recipients (Fig. 4A). Importantly, S. flexneri 2a OPS-IpaB and OPS-CRM elicited antibodies that mediated serum bactericidal activity (SBA) against S. flexneri 2a; SBA titers were .10-fold higher than baseline in both groups (Fig. 4C). In contrast, no SBA activity was detected against S. sonnei Moseley and 53G strains (Fig. S2A). Sera from mice immunized with IpaB alone also lacked bactericidal activity (Fig. S2B). These results are consistent with the notion that OPS-specific antibodies are the main contributors to bactericidal activity and the recognized serotype specificity of OPS-induced immunity. In agreement, mice immunized with sublethal doses of S. tion. Immunization with S. flexneri 2a OPS-IpaB conjugate vaccine afforded 78% protection against homologous S. flexneri 2a challenge (P , 0.0001) whereas S. flexneri 2a OPS-CRM provided only 50% protection (P = 0.0014) (Fig. 5A). IpaB alone conferred 67% protection against S. flexneri 2a (P = 0.0003) (Fig. 5A). The trend of higher protective efficacy of OPS-IpaB than of OPS-CRM or IpaB alone suggests that both OPS and IpaB contribute to the observed protective immunity. IpaB-and OPS-specific IgG titers in mice that were protected were significantly higher than titers in those that succumbed to infection (Fig. 5C).
Importantly, S. flexneri 2a OPS-IpaB exhibited 56% protection against heterologous S. sonnei challenge (P , 0.0001) (Fig. 5B). Because Shigella O-polysaccharide immunity is serotype specific, this cross protection is attributable to IpaB. This is consistent with the lack of protection in the OPS-CRM group (11% survival). IpaB alone afforded 44% protection against S. sonnei (P = 0.0003) (Fig. 5B), which was not significantly different Shigella OPS-IpaB Conjugate Vaccine mSphere from the protection elicited by S. flexneri 2a OPS-IpaB. IpaB-specific serum IgG was again significantly higher in mice that were protected against S. sonnei infection (Fig. 5D). S. flexneri 2a and S. sonnei IpaBs share .98% homology; therefore, cross protection was expected. The slight difference in IpaB efficacy in the two experiments is likely due to the higher severity of S. sonnei infection (mice succumbed sooner). Unvaccinated control mice had very low to no survival. A summary of the protective vaccine efficacy data is shown in Table 1.

DISCUSSION
There is an urgent need for a well-tolerated, effective, and affordable vaccine to prevent Shigella-induced diarrhea. At-risk groups who will benefit the most include children (3 to 36 months of age) living in low-and middle-income countries, travelers to regions of endemicity, military personnel, the elderly, and immunocompromised individuals. We report here for the first time the successful development of a Shigella OPS-conjugate vaccine that employs IpaB as a carrier antigen. IpaB is known to be potently immunogenic (17,18,22,26). IpaB-specific antibodies are present in high levels in sera from individuals living in areas of endemicity (27). Several groups have also reported the protective capacity of IpaB in animal models (17,18,28), and IpaB-specific antibodies (19,20) and memory B cells (29) have been associated with reduced risk of shigellosis in humans.
The OPS-IpaB conjugate described in this study has several important advantages over existing Shigella conjugates: (i) the inclusion of IpaB results in more robust antibacterial immunity that targets the bacterial lipopolysaccharide as well as a virulence factor; (ii) immunity is no longer limited to the serotype-specific O antigen but now expands to a highly conserved antigen; (iii) it avoids the use of carrier proteins already included in several routine vaccines given to infants of the same age, which may interfere with vaccine responses. Additionally, two innovative features distinguish the production of OPS-IpaB from other Shigella conjugate vaccines: (i) efficient production of large amounts of nnAA-containing IpaB using the scalable Xpress 1 CFPS platform-Vaxcyte has scaled CFPS to a 200-L-plus volume under good manufacturing practices (GMP) for production of other carrier proteins-and (ii) the site-specific conjugation of OPS to IpaB, which results in a more predictable and consistent product. IpaB has traditionally been produced in cell-based heterologous expression systems. However, this method has major limitations with regard to recovery and solubility (30)(31)(32). CFPS allows competent production of both native and nnAA-containing soluble IpaB at high yields and low cost. Unlike conventional methods that employ random conjugation, site-specific conjugation preserves functional and/or immunogenic epitopes.  S. flexneri 2a OPS-IpaB administered to mice i.m. in the presence of a traditional alum-based adjuvant was highly immunogenic and afforded high levels of protection against S. flexneri 2a lethal pulmonary infection. Mice that survived the lethal infection had higher levels of both OPS-and IpaB-specific IgG than did those that succumbed to infection, suggesting that both antigenic targets contributed to protection against S. flexneri 2a infection. The OPS-IpaB conjugate elicited OPS-specific antibodies that displayed functional (bactericidal) activity against S. flexneri 2a but not S. sonnei, which confirms the serotype-specific nature of OPS-induced immunity. Serotype-specific SBA responses have been reported in mice immunized i.m. with SF2a-TT15 (33), in rats that received S. flexneri 2a-rEPA (34), and in rabbits vaccinated with a monovalent and quadrivalent Shigella OPS-rEPA bioconjugate (35). The production of antibodies that can activate complement and mediate bacterial killing is important as Shigella SBA has been associated with reduced risk of S. flexneri 2a (19,36) and S. sonnei (37) infection in humans. Further studies to elucidate the breadth of responses elicited by Shigella OPS-IpaB (duration, T cell responses, B memory, cytokine profiles, etc.) are warranted.
The S. flexneri 2a OPS-IpaB conjugate exhibited higher protective efficacy against homologous (S. flexneri 2a) lethal challenge than did OPS-CRM, which can be attributed to immunity against IpaB. This result suggests that a Shigella OPS-IpaB conjugate may provide protection superior to that of other current Shigella OPS conjugate vaccines, which use irrelevant proteins as carriers. Intriguingly, there are no reports in the literature describing the protective efficacy of S. flexneri 2a OPS-rEPA (Flexyn2a), S. sonnei OPS-rEPA, or SF2a-TT15 conjugate vaccines in animal models. Only antibody-mediated blocking of S. flexneri 2a infection in mice, not vaccine protection due to active immunization, was shown for SF2a-TT15 (33). Ours is the first demonstration of significant protective efficacy of a parenterally delivered Shigella conjugate vaccine in a lethal mouse infection model.
Importantly, the S. flexneri 2a OPS-IpaB conjugate also elicited robust cross-protective immunity against S. sonnei. The lack of protection in S. flexneri 2a OPS-CRM-vaccinated mice against a heterologous S. sonnei strain and the serotype-restricted SBA responses elicited by both OPS conjugates confirm that the cross protection afforded by the OPS-IpaB conjugate is dependent not on immunity to OPS but on that to IpaB. Another subunit Shigella vaccine that combines serotype-specific antigens with conserved proteins (IpaB and IpaC) is the Invaplex (38). Cross protection against heterologous serotypes of Shigella has been reported for the Invaplex vaccine in preclinical studies and attributed to the immune response stimulated by IpaB and IpaC (38,39). Unlike sera from OPS-IpaB and OPS-CRM vaccine recipients, sera from mice that received IpaB alone did not exhibit SBA. These results confirm the OPS specificity of bactericidal humoral immunity. The exact mode of action of the IpaB-specific antibodies is not fully understood, although they are presumed to block Shigella host cell invasion. Recent system serology data produced by our group suggest that anti-IpaB antibodies may also facilitate phagocytosis by neutrophils and monocytes (20). Clearly, the inclusion of IpaB as carrier protein not only enhanced but also expanded the conjugate's protective capacity (IpaB homology across Shigella species is .98%). The broader coverage of OPS-IpaB may also simplify the vaccine approach by circumventing the need for multiple conjugates to prevent illness caused by circulating serotypes. Further studies are ongoing to ascertain whether S. flexneri 2a OPS-IpaB or a bivalent S. flexneri 2a and S. sonnei OPS-IpaB conjugate vaccine formulation would be preferable to cover most prevalent Shigella serotypes.
Another advantage of the Shigella OPS-IpaB conjugate vaccine candidate is the use of a carrier protein (i.e., IpaB) that has not been used in other conjugate vaccines given to infants. Carrier proteins in existing Shigella OPS conjugates (e.g., TT and CRM 197 ) are included in multiple other routine conjugate vaccines given to infants of the same age. Immunity to glycoconjugate carriers (preexistent or concurrently induced) has been shown to reduce responses to antigens coadministered with the same carriers. For instance, immunity to a TT-conjugate vaccine was reduced as a result of prior (40) or concomitant (41) immunization with a different TT-containing vaccine-this effect is known as carrier-induced immune suppression. In addition, coadministration of multiple CRM 197 -containing conjugate vaccines is at higher risk of inducing bystander interference than are schedules that do not include CRM 197 (reviewed in references 42 and 43). The use of IpaB in our OPS glycoconjugate avoids these concerns, particularly as new conjugates are added to the infant vaccination schedules. The effectiveness of IpaB on its own, the breadth and features of IpaB immunity, and its properties as a carrier remain to be elucidated.
To the best of our knowledge, we are the first to report the development of a Shigella OPS-IpaB conjugate vaccine candidate and its ability to confer robust protective immunity against multiple Shigella serotypes. There are momentum in the field and strong recommendations for the advancement of Shigella vaccines (44). The OPS-IpaB glycoconjugate candidate described here meets the World Health Organization preferred product profile specifications for vaccines against Shigella (45) in terms of improved antigenic composition, expected safety based on the precedent of other conjugate vaccines, noninterference, parenteral delivery, and potential for robust immunogenicity and broad efficacy; its holds great potential for prevention of disease in the most vulnerable groups (i.e., infants and toddlers). Efforts to expand product analytical characterization and preclinical studies of immune responses and efficacy against other Shigella serotypes and clinically relevant strains with eventual evaluation in humans are ongoing.

MATERIALS AND METHODS
Production of Shigella core OPS. S. flexneri 2a core OPS production was performed as previously described for purification of Salmonella enterica serovar Typhimurium core OPS (46), with modifications. Briefly, the core OPS was extracted from a live attenuated S. flexneri 2a strain (CVD1204) (21). The bacteria were grown in a chemically defined medium supplemented with 0.13 g/L amino acids, 0.025% guanine, and 50 mg/mL kanamycin to stationary phase, and core OPS was extracted by boiling the fermentation culture in ;6% acetic acid, pH 3.7, at 100°C for 4 h. Cellular debris was removed by centrifugation at 10,000 Â g, 4°C, for 30 min, and the supernatant was brought to 1 M NaCl and clarified using a 0.2-mm hollow-fiber tangential flow filtration (TFF) membrane (Sartorius, Germany). The OPS in the filtered supernatant was concentrated over a 30-kDa Hydrosart TFF membrane and diafiltered 35-fold with 1 M NaCl to remove impurities, followed by 10-fold diafiltration with 20 mM Tris, 50 mM NaCl, pH 7, for buffer exchange into appropriate buffer for anion-exchange chromatography. The concentrated, bufferexchanged OPS was passed over a Sartobind Q membrane (Sartorius, Germany) in negative mode to remove impurities, and the OPS was collected in the flowthrough. The OPS was then brought to 25% ammonium sulfate and incubated at 4°C for ;20 h. Precipitate was removed by centrifugation at 10,000 Â g, 4°C, 30 min, followed by 0.45-mm filtration. The core OPS in the filtered supernatant was concentrated using a 10-kDa Hydrosart TFF membrane and diafiltered 10-fold into water. The core OPS was sterile filtered with a 0.2-mm vacuum filter, frozen, and lyophilized for dry-weight analysis. Identity for sugar moieties in the purified OPS was confirmed by high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD) with an ICS-6000 system and CarboPac PA10 column run at 0.35 mL/min using commercially available monosaccharide standards.
Cloning, expression, and purification of nnAA-containing variant of IpaB. The genes for expression of IpgC (amino acids [aa] 1 to 155, GenBank accession number AAP78992.1) and native or pAMFcontaining IpaB (aa 1 to 580, GenBank accession number SVH88885.1) from S. flexneri 2a were synthesized at ATUM (Menlo Park, CA) and subcloned with an N-terminal methionine into a proprietary vector using NdeI and SalI as restriction sites. Each of the genes contained a C-terminal His 6 tag for purification of the protein. In vitro native and pAMF-containing protein expression and titer estimation using the XpressCF 1 CFPS platform were performed as described elsewhere (22). For purification of IpaB-variant 1 (expressed in the presence of exogenously added chaperone IpgC at an 0.3-mg/mL concentration), the process was scaled up at a 2-to 4-L reaction volume and controlled CFPS in a bioreactor was performed using a DASbox at 25°C for 10 h with constant stirring at 650 rpm while maintaining the pH at 7.2 and sparging with a blend of air and oxygen through the reaction mixture to maintain the dissolved oxygen at 30%. After 10 h, the reaction mixtures were harvested and spun down at 15,000 Â g at 4°C for 30 min followed by filtration using an 0.45-mm pore size membrane. Thereafter, the clarified filtrate was loaded onto a 5-mL HisTrap Excel column preequilibrated with buffer A1 (50 mM Tris, 150 mM M NaCl, 10 mM imidazole, 0.1% lauryldimethylamine oxide [LDAO]), which in turn also helped remove coeluting IpgC from IpaB. Finally, the bound protein was eluted using a 50% step gradient of buffer A1 with 500 mM imidazole. Postcapture, the elution fractions for IpaB were combined, concentrated using 30-kDa-cutoff Amicon Ultra-15 centrifugal filters (Millipore Sigma, USA), and loaded onto a gel filtration Superdex200 26/60 column preequilibrated with buffer S1 (50 mM Tris, 150 mM M NaCl, 0.1% LDAO). Fractions with the highest purity (.95% as assessed by SimplyBlue SafeStain [ThermoFisher, USA]) after SDS-PAGE analysis were combined, aliquoted, and stored at 280°C for further use. Protein concentration was measured by A 280 while subtracting the background absorbance for the buffer alone. Purified IpaB-var1 was incubated with excess DBCO-TAMRA dye for 1 h at room temperature to confirm pAMF incorporation. SDS-PAGE gel and fluorescence readout was recorded using a Syngene G-box gel imager.
DBCO derivatization of OPS and conjugation to IpaB-var1 for generating conjugate vaccine. The purified OPS resuspended to 6 mM in 100 mM borate buffer (pH 8.5) was combined with 3 equivalents (to the polysaccharide repeating unit [PSRU]) of 1-cyano-4-dimethylaminopyridinium tetrafluoroborate (CDAP; from 100-mg/mL solution in acetonitrile) with vigorous stirring to facilitate cyanylation at reactive hydroxyl groups. At 5 min after the addition of CDAP, 2 molar equivalents of dibenzocyclooctyne-amine linker stock in dimethyl sulfoxide (DMSO) was added to reach a final DMSO concentration of 5% (vol/vol). After DBCO derivatization, unreacted cyanate esters were quenched by addition of 200 mM glycine. The DBCO-derivatized OPS (OPS-DBCO) was purified via a Zeba spin column, and the purity of the material was assessed by reverse phase. A single peak in high-performance liquid chromatography (HPLC) when absorbance was monitored at 309 nm confirmed the complete removal of excess DBCO linker and other reaction by-products. Finally, the OPS concentration was measured using an anthrone assay, and dibenzocyclooctyne concentration was measured using absorbance at 309 nm. These two values were combined to give an estimate of the percentage of polysaccharide derivatized with a dibenzocyclooctyne functional group. For conjugation, the percent DBCO derivatization of the mutant polysaccharide (PS) was kept between 5 and 10%. Thereafter, IpaB-var1 was mixed with OPS-DBCO at a 1:1 ratio (0.5 mg/mL of each) to facilitate conjugation via click-chemistry reaction at room temperature. Postreaction, the reaction mixture was dialyzed against a 50-kDa-cutoff membrane to remove excess unreacted OPS-DBCO. The recovered conjugates were analyzed by SEC-MALS, and the final concentration of OPS and proteins in the purified conjugates were measured using anthrone and bicinchoninic acid (BCA) (Pierce protein assay kit; Thermo Scientific) assays, respectively. Both OPS and IpaB/CRM were present at a 1:1 ratio in the final purified conjugates; the 10-mg immunizing doses contained 10 mg of OPS conjugated to 10 mg of protein.
SEC-MALS analysis. The SEC-MALS UV-RI setup consists of an Agilent HPLC 1100 degasser, temperature-controlled autosampler (4°C), column compartment (25°C), and UV-visible (UV-Vis) diode array detector (Agilent, Santa Clara, CA) in line with a Dawn-Heleos multiangle laser light scattering detector and Optilab T-rEX differential refractive interferometer (Wyatt Technology, Santa Barbara, CA). The system was coupled to a Superdex200 10/30 Increase column for IpaB-var1 protein. A mobile phase consisting of 0.2-mm-filtered 50 mM Tris (pH 8), 150 mM NaCl, 0.1% (vol/vol) LDAO was used at a 0.5-mL/min flow rate. Approximately 50 to 100 mg of sample was injected for analysis. Agilent Open Lab software was used to control the HPLC, and Wyatt Astra 7 software (Wyatt Technology Corp., Santa Barbara, CA) was used for data collection and molecular weight analysis.
Anthrone assay for estimating total polysaccharide concentration. A stock of 2 mg/mL of the anthrone reagent (Sigma-Aldrich, St. Louis, MO; CAS number 90-44-8) was prepared in cold sulfuric acid while a 1 mM stock of polysaccharide repeating unit (PSRU) was prepared in water as a standard. In triplicate wells, 100 mL of PSRU stock (serially diluted into reference standards) or the unknown samples (diluted 1:3) was plated (96-well plate) followed by addition of 200 mL/well of the anthrone reagent stock. All reaction mixtures were thoroughly mixed and sealed with a plate cover for incubation at 95°C for 10 min. The plate was briefly placed on ice to cool to ambient temperature before absorbance was measured at 620 nm using a UV-Vis plate reader. To determine concentration of unknown samples, PSRU standard concentrations and absorbances were used to generate a leastsquare fit regression.
Ethics statement. The current study was performed in accordance with guidelines from the "Guide for the care and use of laboratory animals" of the National Institutes of Health (NIH). All animal studies and procedures were approved by the University of Maryland School of Medicine Institutional Animal Care and Use Committee (IACUC). All relevant ethical regulations for animal testing and research were followed. All efforts were made to minimize pain and distress of the animals.
Mice, immunizations, and challenge. Adult female BALB/c mice (6 to 8 weeks old) were purchased from Charles River Laboratories, Wilmington, MA. Mice were immunized i.m. on days 0 and 28 with 10 mg S. flexneri 2a OPS-IpaB, S. flexneri 2a OPS-CRM, and IpaB adsorbed to AdjuPhos (4.8%, vol/vol; InvivoGen, CA). The OPS-IpaB or OPS-CRM doses contained 10 mg of each antigen in the conjugate. i.m. immunization was performed by dispensing a 100-mL volume (50 mL per leg). Negative-control mice received i.m. AdjuPhos or PBS. Positive-control groups were immunized i.n. on days 0 and 28 with sublethal doses (;1 Â 10 5 CFU) of S. flexneri 2a 2457T or S. sonnei Moseley for respective challenge strains. At 4 weeks after the last immunization, mice were challenged i.n. with ;5 Â 10 6 CFU of S. flexneri 2a 2457T or S. sonnei Moseley (corresponding to ;4 or ;2 50% lethal doses [LD 50 ], respectively). S. flexneri 2a and S. sonnei inoculum preparation and i.n. challenge were performed as previously described (17). Mice were monitored daily for 15 days after the challenge, and their health status, weight, and survival were recorded. Mice that reached a moribund state or lost more than 20% of their initial body weight and did not recover within 72 h were humanely euthanized and considered nonsurvivors. Blood was collected at day 21 (prevaccination) and days 13, 27, 42, and 55 (postvaccination) for serum antibody measurement. Blood collection via retro-orbital or submandibular vein and i.n. infection were performed under isoflurane anesthesia (VetEquip, Inc., CA).
Antibody measurement. Mouse IpaB-, S. flexneri 2a OPS-, and CRM-specific serum antibodies were measured by enzyme-linked immunosorbent assay (ELISA) as previously described (17,22). For measurement of conjugate vaccine-induced OPS-specific antibodies, Nunc MaxiSorp plates (Thermo Scientific) were coated with S. flexneri 2a OPS-CRM (5 mg/mL) or OPS-IpaB (5 mg/mL) conjugates; for analysis of IpaB-or CRM-specific antibodies, Immulon 2HB plates (Thermo Scientific) were coated with IpaB (0.1 mg/ mL) or CRM (5 mg/mL), respectively. Serum samples were tested in 2-fold serial dilution in duplicate. Horseradish peroxidase (HRP)-labeled goat anti-mouse IgG (KPL SeraCare, Gaithersburg, MD) was used as detection antibody. For evaluation of IpaB immune reactivity, plates were coated with IpaB or IpaB variants (produced by CFPS). A pool of human convalescent-phase sera was added and serially diluted 2-fold starting at 1:20,000. Bound antibodies were detected with HRP-goat anti-human IgG (Jackson Immuno Research, West Grove, PA). Both values of optical density at 450 nm (OD 450 ) values and endpoint titers were obtained. Endpoint titers were calculated by interpolation of absorbance values of samples in the linear regression curve of a calibrated in-house standard and were reported as ELISA units (EU)/mL. An ELISA unit corresponds to the reciprocal serum dilution resulting in an absorbance value of 0.2 at 450 nm above the background.
SBA. The serum bactericidal activity (SBA) assay was conducted as previously described (47), with modifications. Briefly, heat-inactivated mouse sera were serially diluted 3-fold in assay buffer in 96-well U-bottom plates (Fisher Scientific). S. flexneri 2a 2457T or S. sonnei Moseley and 53G strains (;120 CFU in 10 mL) and 50 mL of 20% baby rabbit complement (BRC) (Pel-Freez Biologicals, AR) were then added to the wells. Plates were incubated for 2 h at 37°C without shaking. Negative controls contained the bacteria and BRC only. A standard with assigned SBA titer was included in each assay and used for titer normalization. Colony counts were enumerated using NIST Integrated Colony Enumerator (NICE) software, and an Excel-based software program, Opsotiter, was used to determine SBA titers.
Statistical analysis. Statistical analyses were conducted using GraphPad Prism 9 (GraphPad Software, La Jolla, CA). Antibody titers were analyzed by unpaired t test. Survival curves of vaccinated and control groups were analyzed and compared using a log rank (Mantel-Cox) test. P values of ,0.05 were considered statistically significant.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.

ACKNOWLEDGMENTS
The study was supported by Vaxcyte through a laboratory service agreement with the University of Maryland.
We thank Sharon Tennant and the CVD Protein Purification Unit for providing strains and reagents, and Dylan Dunn and Emily Levesque for technical assistance.