2c and a), in contrast to what was obtained with NaIO4

(Fig. 2b). OAg-oxTEMPO Protein Tyrosine Kinase inhibitor with an average percentage number of oxidized repeating units of 36% and 15% were conjugated to CRM197, to investigate the impact of the degree of OAg derivatization on the immunogenicity of the corresponding conjugates. The same conditions for the conjugation and purification of OAg-oxNaIO4 were applied and in both cases all CRM197 in the reaction mixtures was conjugated, with 19–28% of OAg conjugated (Fig. 3b). Conjugates obtained using less derivatized OAg (both after treatment with NaIO4 or TEMPO) were characterized by a higher OAg to protein ratio with respect to the conjugate obtained from more oxidized OAg which was able to couple to more CRM197 molecules (Table 1). The terminal KDO residue of the core oligosaccharide was used for selective linking of OAg to CRM197 without modifying the OAg chain. To generate one conjugate vaccine, reductive amination Cilengitide datasheet with ADH was followed by reaction with SIDEA and conjugation to CRM197[28]. A similar chemistry was evaluated where the first

step of reductive amination was conducted with NH4OAc, allowing the synthesis of a conjugate with a linker about half the length of ADH-SIDEA (Fig. 1b). After testing the reactivity of OAg-KDO with NH4OAc under different conditions (see SI), in order to synthesize the corresponding conjugate, the reaction was performed at pH 7.0 for 5 days resulting in the activation of 90% of OAg chains. Use of the longer ADH linker with the hydrazide functionality allowed Mannose-binding protein-associated serine protease the reaction to proceed, with activation close to 100% after only 2 h at pH 4.5. In the following step where the OAg derivatives were reacted with SIDEA, >90% of total NH2

groups were coupled to SIDEA, for both OAg-NH2 and OAg-ADH. The analysis of the corresponding conjugation mixtures by HPLC-SEC, confirmed conjugate formation without residual free protein, while the amount of conjugated OAg was close to 15% in both cases. The resulting conjugates were very similar in terms of OAg to CRM197 ratio (4–5 OAg chains linked per protein) and molecular size, measured as distribution coefficient Kd by HPLC-SEC; even if OAg-NH2-SIDEA-CRM197 showed a slightly broader population (Table 1, Fig. 3c). Selective conjugates contained higher OAg to protein ratios than random conjugates (Table 1). The synthesized conjugates were tested in mice, with the following main objectives: to compare the immunogenicity of random versus selective conjugates; to analyze the impact of linker chain length on the immunogenicity of selective conjugates; to evaluate whether the degree of random modification of the OAg chain impacts on immunogenicity. After three doses, all the conjugates generated anti-OAg IgG levels that were not statistically different (Fig. 4a).