Sialosides on N- and O-linked glycoproteins play a fundamental role in many biological processes and synthetic glycan probes have proven to be valuable tools for elucidating these functions. Though sialic acids are typically found α2-3 or α2-6-linked to a terminal non-reducing end galactose, poly-LacNAc extended core-3 O-linked glycans isolated from rat salivary glands and human colonic mucins have been reported to contain multiple internal Neu5Acα2-6Gal epitopes. Here, we have developed an efficient approach for the synthesis of a library of N- and O-linked glycans with multi-sialylated poly-LacNAc extensions, including naturally occurring multi-sialylated core-3 O-linked glycans. We have found that a recombinant α2-6 sialyltransferase from Photobacterium damsela (Pd2,6ST) exhibits unique regioselectivity and is able to sialylate internal galactose residues in poly-LacNAc extended glycans which was confirmed by MS/MS analysis. Using a glycan microarray displaying this library, we found that Neu5Acα2-6Gal specific influenza virus hemagglutinins, siglecs and plant lectins are largely unaffected by adjacent internal sialylation, and in several cases the internal sialic acids are recognized as ligands. Polyclonal IgY antibodies specific for internal sialoside epitopes were elicited in inoculated chickens.

Vaccines targeting glycan structures at the surface of pathogenic microbes must overcome the inherent T cell-independent nature of immune responses against glycans. Carbohydrate conjugate vaccines achieve this by coupling bacterial polysaccharides to a carrier protein that recruits heterologous CD4 T cells to help B cell maturation. Yet they most often produce low- to medium-affinity immune responses of limited duration in immunologically fit individuals and disappointing results in the elderly and immunocompromised patients. Here, we hypothesized that these limitations result from suboptimal T cell help. To produce the next generation of more efficacious conjugate vaccines, we have explored a synthetic design aimed at focusing both B cell and T cell recognition to a single short glycan displayed at the surface of a virus-like particle. We tested and established the proof of concept of this approach for 2 serotypes of Streptococcus pneumoniae. In both cases, these vaccines elicited serotype-specific, protective, and long-lasting IgG antibodies of nanomolar affinity against the target glycans in mice. We further identified a requirement for CD4 T cells in the anti-glycan antibody response. Our findings establish the design principles for improved glycan conjugate vaccines. We surmise that the same approach can be used for any microbial glycan of interest.

Heparin is widely used for anticoagulation, often requiring the subsequent administration of a reversal agent. The only approved reversal agent for heparin is protamine sulfate, which induces well described adverse reactions in patients. Previously we reported a novel class of heparin antagonists based on the bacteriophage Qβ platform, displaying polyvalent cationic motifs which bind with high affinity to heparin. Here we report heparin reversal by the most effective of these virus-like particles (VLP) in samples from patients who were administered heparin during cardiac procedures or therapeutically for treatment of various thrombotic conditions. The VLP consistently reversed heparin in these samples, including those from patients that received high doses of heparin, with greater efficiency than a negative control VLP and with significantly less variability than protamine sulfate. These results provide the first step towards validation of heparin antagonist VLPs as viable alternatives to protamine.

The anticoagulant activity of heparin administered during medical interventions must be reversed to restore normal clotting, typically by titrating with protamine. Given the acute toxicity associated with protamine, we endeavored to generate safer heparin antagonists by engineering bacteriophage Qβ virus-like particles (VLPs) to display motifs that bind heparin. A particle bearing a single amino acid change from wild-type (T18R) was identified as a promising candidate for heparin antagonism. Surface potential maps generated through molecular modeling reveal that the T18R mutation adds synergistically to adjacent positive charges on the particle surface, resulting in a large solvent-accessible cationic region that is replicated 180 times over the capsid. Chromatography using a heparin-sepharose column confirmed a strong interaction between heparin and the T18R particle. Binding studies using fluorescein-labeled heparin (HepFL) resulted in a concentration-dependent change in fluorescence intensity, which could be perturbed by the addition of unlabeled heparin. Analysis of the fluorescence data yielded a dissociation constant of approximately 1 nM and a 1:1 binding stoichiometry for HepFL:VLP. Dynamic light scattering (DLS) experiments suggested that T18R forms discrete complexes with heparin when the VLP:heparin molar ratios are equivalent, and in vitro clotting assays confirmed the 1:1 binding stoichiometry as full antagonism of heparin is achieved. Biolayer interferometry and backscattering interferometry corroborated the strong interaction of T18R with heparin, yielding K ∼ 1-10 nM. These biophysical measurements further validated T18R, and VLPs in general, for potential clinical use as effective, nontoxic heparin antagonists.

No abstract