Surface polysaccharides (PS) of pathogenic bacteria, including capsusalar PSs (CPS) or lipopolysaccharides (LPS), serve as both essential virulence factors and protective antigens. The age-related and T cell–independent immunogenicity of CPS limit their use as vaccines, especially in infants and young children. LPS are too toxic to be administered. Accordingly, their O-specific PSs (O-SP), which shares the virulence-promoting and protectiveness properties of CPS, must be purified. However, O-SP are too small to be immunogenic (haptens). Covalent binding of CPS or of O-SP to medically useful proteins to form conjugates both increases their immunogenicity and confers T-cell dependence on these saccharides. Part of the work of this laboratory is directed at making synthetic vaccines to improve their immunogenicity.

Conjugate Vaccines against Shigellosis
The O-SP of Shigella sonnei and of Shigella flexneri 2a were bound to bacterial toxoids. In adult army recruits and then in four- to seven-year-old children, both conjugates proved to be safe and induced statistically significant and long-lived rises in IgG antibody levels against the homologous LPS. Similar though lesser increases in IgM and IgA anti–LPS levels were also induced. Re-injection of S. flexneri 2a conjugate induced a booster response in the recruits as well as in the four- to seven-year-old children. A Phase III trial showed that one injection of S. sonnei O-SP, bound to a nontoxic recombinant Pseudomonas aeruginosa exoprotein A (rEPA), protected army recruits against outbreaks of this pathogen. Importantly, we noted a statistically significant correlation between the levels of serum IgG anti–LPS and the efficacy of the conjugate. We developed two methods that increased the immunogenicity of the Shigella conjugates in mice: another carrier protein, a genetically inactivated Corynebacterium diphtheriae toxin (CRM9), was a superior carrier for S. sonnei O-SP, and treatment of rEPA with succinic anhydride, a nontoxic mild akylating agent that converts amino groups of proteins to carboxyls, increased the immuno-genicity of S. flexneri 2a O-SP. A Phase I study of these Shigella conjugates in adults confirmed their safety and immunogenicity; the improved immunogenicity was less marked than in mice. A Phase II study in one- to four-year-old children showed an improved immunogenicity of the new S. flexneri 2a conjugate but less immunogenicity of the S. sonnei conjugate. A Phase III study of the modified S. flexneri 2a and the original S. sonnei conjugates is in preparation. In collaboration with the Lanzhou Vaccine Institute and Provincial Medical Center in Henan, China, a clinical trial of the two conjugates is in the planning stage.

Conjugate Vaccine against Staphylococcus aureus
Schneerson, Robbins; in collaboration with Shinefeld, Black, Fattom, Horwith, Fuller, Ordonez, Rasgon, Yoh, Law, Johnson, Alcom, Muenz, Naso
Staphylococcus aureus, an opportunistic pathogen, is a major cause of bacteremia and its complications in patients with decreased resistance. The pathogen is becoming increas-ingly resistant to last-resort antibiotics such as methicillin and vancomycin. Its CPS is essential for pathogenesis of and immunity to S. aureus. In a randomized, double-blinded, placebo-controlled study conducted in about 1,800 hemodialysis patients, we found that our conjugate vaccine consisting of the CPS linked to recombinant Pseudomonas aerugi-nosa nontoxic exotoxin A significantly decreased the number of S. aureus infections. Given that such patients are least likely to respond to immunoprophylaxis, the vaccine’s efficacy is likely to be higher in other patients at risk for S. aureus infection.

Recombinant Vaccine against Bacillus anthracis
Schneerson; in collaboration with Leppla, Ramirez, Shiloach
The virulence of Bacillus anthracis is mediated by a capsule and an exotoxin consisting of protective antigen (PA), lethal factor, and edema factor. Antibodies against PA are thought to be protective. The currently available vaccine, produced by treating B. anthracis culture supernatants with formalin, has a relatively high rate of side effects and has to be administered several times to elicit protective immunity. Moreover, because it consists of a mixture of undefined components, it is difficult to standardize. To overcome these problems, we developed culture conditions and purification procedures to produce sufficient quantities of a genetically modified PA (rPA) to test in mice. The resultant rPA vaccine elicited high levels of IgG anti-PA antibodies with neutralizing activities in mice.

Use of Cross-Reacting Antigens to Boost Immungenicity
To determine if concurrent administration of a cross-reacting and homologous CPS has an advantage over the use of the homologous CPS alone, we used conventional methods to isolate the cell wall PS of Bacillus pumilus, SH18, which has been reported to cross-react with the CPS of Haemophilus influenzae type b (HIb). We then investigated its structure by using gel chromatography–mass spectrometry and showed that it contains glycerol, ribitol, and 2-acetamido-2-deoxyglucose, in a molar ratio of 0.2:1.0:0.2 and 17 percent phosphate. The CPS cross-reacted with anti–Staphilococcus epidermidis as well as with the anti-HIb. Methods to prepare a conjugate of this PS are under investigation.

Other Vaccines
Neisseria meningitidis group A causes endemic and epidemic meningitis, notably in the meningitis belt of Africa. An effective and available CPS vaccine is underused. To improve the vaccine’s immunogenicity further, as in the case of other CPS, we are investigating methods of binding it to a carrier protein. We developed a double mutant of Bordetella pertussis, producing a genetically inactivated toxin deficient in FHA synthesis. Effort is currently directed toward increasing production of this B. pertussis strain as a more easily purified pertussis toxin for a monocomponent vaccine and as a carrier protein for pneumococcal type 14 CPS. Clostridium difficile is a major cause of hospital-acquired diarrhea following antibiotic usage: the diarrhea caused by the pathogen is mediated by two exotoxins, A and B. Toxin A, considered to be the major toxin, can, in extreme cases, cause pseudomembranous colitis. A genetically derived toxin mutant (rARU) both induces antitoxin and protects animals from infection with C. difficile. The succinylation of rARU improved its solubility and did not detectably affect its antigenicity. Techniques to prepare a mutant toxin A for clinical use have been worked out. We conjugated three PSs of varying composition from pneumococcus type 14, Escherichia coli K1, and S. flexneri 2a to succinylated rARU. The resultant conjugates induced high levels of both antiPS and antitoxin. Preparation of toxin A conjugates for clinical evaluation is under way. Borrelia burgdorferi, a spirochete transmitted though the bite of infected Ixodes ticks, is the etiologic agent of Lyme disease. A protein vaccine against it is available but is not effective under the age of 12 years. Lipopolysaccharide (LPS) has been described in other spirochetes, but its presence in B. burgdorferi has been debated. So far, we have not been able to confirm its presence. The search for LPS revealed a unique glycolipid consisting of glycerol and galactose as the carbohydrate moiety. Evidence suggests that this glycolipid is surface-exposed. Injected in complete Freund’s adjuvant, it induced specific antibodies.

Synthetic Vaccines against Shigella dysenteriae Type 1 and Shigella sonnei
Ekborg, Sampath-Kumar, Nelson, Pozsgay, Robbins, Schneerson; in collaboration with Liptak
Shigella dysenteriae type 1 and Shigella sonnei are Gram-negative human pathogens that cause endemic and epidemic dysentery throughout the world. Despite the pathogens’ discovery many years ago, no licensed vaccines exist to protect against the bacteria, which have developed resistance to most available antibiotics. Our approach to vaccine development is based on the demonstration that serum antibodies against the O-SPs of Gram-negative bacteria are important for host protection. While O-SPs are nonimmungenic, presumably because of their low molecular weight, covalent conjugates of an immuno-genic protein and the O-SP of S. dysenteriae type 1 and S. sonnei have been shown to elicit significant anti–O-SP antibody levels of the IgG isotype, which may be boosted by repeated injections. We surmised that an improved vaccine might be constructed from chemically defined oligosaccharide fragments of O-SPs, which are devoid of biological contamination, carry a uniform molecular weight, and can be characterized by physico-chemical methods.

We have chemically synthesized oligosaccha-ride fragments of the native O-SP of S. dysenteriae type 1 up to a tetracosasaccharide consisting of 24 monosaccharide residues. These constructs are attached to a spacer that makes bioconjugation possible through the termini of the spacer moieties (i.e., at a single site). We have been developing a new technique for chemical synthesis of extended oligosaccharides that circumvents the difficulties associated with the conventional, liquid-phase coupling of oligosaccharide building blocks and avoids the problems inherent in the solid-phase approaches. In brief, our method employs lipophilic protecting groups in one of the blocks while using conventional protecting groups in the others. The presence of the lipophilic groups in the targeted intermediate but not in the side-products allows the isolation of the products by the solid-phase extraction technique that uses recyclable chromatographic materials and environmentally friendly alcoholic solvents. To improve the efficiency of oligosaccharide synthesis by facilitating product isolation, we have designed and synthesized several new lipophilic protecting groups, including 4-dodecyloxybenzyl, 4dodecylbenzyl, and 4-octadecylbenzyl moieties. The utility of these groups has been documented. We are currently using our new technology for the synthesis of the oligosaccharides related to the O-SP of S. dysenteriae type 1 under clean laboratory conditions for Phase I clinical trials. So far, we have demonstrated that the immunogeni-city of the synthetic saccharide–human serum albumin conjugates depends on the saccharide chain length and the saccharide density on the protein, the latter in a nonlinear fashion.

Current work is directed at determining the role, if any, of the monosaccharide terminating the oligosaccharide chain. To this end, we have synthesized oligosaccharide fragments of the OSP that differ in the terminal unit, i.e., in the repeating unit frame. We designed experiments aimed at making a PS mimic that might be more immunogenic when conjugated to proteins than the saccharides synthesized so far. In brief, we designed linearly arranged “clusters” of the synthetic oligosaccharides that are intercon-nected by spacers through predefined sites. We demonstrated the feasibility of this approach by interconnecting six monosaccharide units. Next, we extended this approach to a synthetic octasaccharide. So far, we have succeeded in interconnecting three such units, leading to a polymer containing 24 monosaccharide residues. At present, we are trying to synthesize even longer constructs as protein conjugates for immunological experiments.

We developed a new technique for reagentless attachment of oligosaccharides to proteins. In short, the technique is based on the Diels-Alder cycloaddition between an activated double bond and a conjugated diene. The activated double bond component was introduced in the protein by using the commercial reagent 3sulfo-succinimidyl 4-maleimidobutyrate. The carbo-hydrate sector was equipped with various linkers containing a 1,3 conjugated diene system. When solutions of the functionalized components were combined, conjugation took place at room temperature. pH has a marked effect on the efficiency of the coupling: lower pH (e.g., pH 5.7) favors conjugation compared with higher pH (e.g., 9.5). The conjugation was as effective in water as in pH 5.7 buffer. On the other hand, the structure of the diene system has little effect on the conjugation reaction. An advantage of this method over existing methods is that the unconjugated saccharide can be recovered with its linker for reuse in a subsequent conjugation experiment.

Synthetic Vaccines against Group A Neisseria meningitidis
Berkin, Coxon, Yergey, Pozsgay, Robbins, Schneerson
Group A Neisseria meningitidis causes endemic and epidemic meningococcal meningitis worldwide. Although a vaccine containing the purified CPS of the bacterium has been available for years, it is not sufficiently immunogenic in infants, who are at highest risk. Our program to develop a more efficient vaccine is based on the assumption that protein conjugates of relatively short fragments of the CPS (i.e., which can be made available by controlled chemical synthesis) of Group A N. meningitidis will be more immunogenic in infants than the currently available vaccine. The CPS of Group A N. meningitidis consists of alpha-linked N-acetyl-mannosaminyl residues that are interconnected by an anomerically located phosphodiester linkage. The native PS is nonstoichiometrically O-acetylated at the O3 position. In our first approach to studying the requirements of synthetic oligosaccharides related to this polymer, we targeted the synthesis of non–O-acetylated congeners. We have designed a route to the mannosaminyl H-phosphonate moiety. Iterative condensation of the unit allowed the synthesis of a dimer and trimer of the PS’s repeating unit in bioconjugatable form. We also synthesized the mannosaminyl spacer and its phosphate ester. Covalent attachment of the synthetic subunits to human serum albumin using our conjugation technology afforded glycoconju-gates containing up to 30 saccharide chains per molecule of HSA. The antigenicity of the conjugates was demonstrated by the precipitin formation with the serum of a horse that was immunized with formalin-killed Group A N. meningitidis. The experiments showed that the serum recognizes as small a component of the PS as a monosaccharide. The formation of precipitation indicates that the presence of O-acetyl groups in the polymer is not essential for antigenicity. Current work is directed toward the synthesis of higher oligomers with and without O-acetyl groups for evaluation of their immuno-genicities.