Robert J. Woods

Linking genetic variation to functional phenotype remains a major barrier to assessing the cross-host potential of emerging viruses. Here, we reconstruct the evolution of predicted hemagglutinin (HA) phenotypic traits across ~13,000 highly pathogenic avian influenza A H5N1 clade 2.3.4.4b viruses circulating in North America. The wide geographic spread of avian influenza within North America triggered a wave of broad HA phenotypic diversity that was later refined by the selective sweeps in avian hosts. Following establishment in dairy cattle, however, viral populations exhibited renewed phenotypic diversification in HA, including increased permissiveness for α2,6-linked sialoside engagement despite high conservation of the receptor-binding domain. These patterns indicate that cattle-associated HA phenotypes can draw from standing predicted HA phenotypic breadth within circulating viral populations while continuing to diversify during cattle-associated circulation, rather than following only a simple stepwise adaptive path through canonical receptor-binding substitutions. By linking viral sequence variation to predicted protein properties across naturally evolving populations, this framework provides a scalable strategy for prioritizing H5N1 variants with cross-host-relevant features for targeted surveillance and experimental follow-up.

We present online three-dimensional (3D) structure-prediction tools at GLYCAM-Web ( www.glycam.org ) that can be used for generating experimentally consistent 3D structures of oligosaccharides for data interpretation, hypothesis generation, 3D visualization, molecular docking and further simulation. The tools support the modeling of an unlimited array of natural glycans and polysaccharides, glycosaminoglycans, engineered glycomaterials and glycoproteins. GLYCAM-Web is directly linked to external databases, such as the Protein Data Bank, facilitating comparison with experimental data.

Exopolysaccharides are carbohydrate polymers secreted by bacteria to perform various roles including adhesion to surfaces, protection from environmental stressors, and as key components in the production of biofilms. The human pathogen Vibrio cholerae produces an exopolysaccharide called VPS, which is essential for the formation of its biofilm matrix through crosslinking interactions with a series of secreted accessory proteins. VPS consists of a repeating tetrasaccharide unit consisting of a uniquely modified α-L-gulose moiety (with N-acetyl, O-acetyl, and amide-linked glycine modifications). Encoded within the cluster containing the biofilm-production genes is a glycoside lyase called RbmB, which cleaves VPS and has been implicated in biofilm dispersal. Here we describe the ~2 Å X-ray structure of RbmB bound to tetrameric and octameric fragments of VPS representing 10 total monosaccharides of an enzyme–product complex. The structure, along with molecular dynamics simulations and complementary functional analyses of mutant proteins in vitro and in vivo, illustrates how recognition and cleavage of VPS relies on salt-bridging interactions between arginine residues in RbmB and the amide-linked glycine moiety uniquely found on the VPS L-gulose and a ~30° bend near the cleavage site that strains the scissile bond. We further present results regarding the localization of RbmB in V. cholerae cells, suggesting that cleavage of VPS by RbmB likely takes place in the periplasm. Sequence conservation suggests that homologous biofilm systems found in many Vibrio species likely utilize a similar VPS cleavage mechanism suggesting that RbmB could be an effective tool for dispersing biofilms across the Vibrio genus and beyond.

Here we present the GLYCAM Bacterial Carbohydrate Builder ( https://glycam.org/cb ), an enhanced version of the GLYCAM-Web Carbohydrate Builder ¹ structure modeller that integrates support for modelling bacterial glycans, enabling the straightforward generation of three-dimensional structural models. The tool integrates bacterial monosaccharide parametrisations into a curated, user-friendly web-based resource. It provides an intuitive interface for the generation of carbohydrate sequences and generates 3D structural models in PDB file format, as well as the input files required for performing molecular dynamics simulations with the AMBER software package. The current implementation includes a library of 18 bacterial monosaccharides, which can be used in combination with the already parametrised eukaryotic sugars to construct complex bacterial glycans. Common derivatives, including acetylation, methylation, and sulfation are also supported. By validating and integrating bacterial sugar parameters into the GLYCAM-Web Carbohydrate Builder, this work reduces the technical barriers associated with bacterial glycan modelling and facilitates computational studies of complex bacterial glycoconjugates.

Engineered glycomaterials represent an exciting new field of biomaterials, owing to their vast structural diversity, yielding a myriad of potential properties and applications. Glycomaterials can be composed of naturally occurring polysaccharides (cellulose, hyaluronic acid, chondroitin sulfate, etc.), but these are also amenable to chemical derivatization, resulting in engineered glycomaterials with altered chemical and material properties. However, rules for predicting the properties of glycomaterials, based on their chemical structure, are not well established, hindering their rational design. Computational methods, such as molecular dynamics (MD) simulation, can accurately characterize the spatial and temporal properties, of glycomaterials; however, the application of MD simulations to predict material properties, such as diffusion, solubility, viscosity, and hydrogel formation, has received less attention. This work demonstrates that diffusion properties of well-known glycomaterial constituents, measured by DOSY NMR spectroscopy and calculated from explicit solvent MD simulations with the GLYCAM06 force field, generally agree well. However, the theoretical results are found to be heavily dependent on the water model, with the TIP5P and OPC models outperforming the widely used TIP3P model. Lastly, an empirical method for estimating the diffusion properties of carbohydrates, based on assessing the number of tightly bound waters, is proposed. Together, these results illustrate the potential of computational approaches to guide the rational design of engineered glycomaterials.

We describe a machine-learned (ML) model, MCNet, which predicts interactions between proteins and glycans. MCNet predicted quantitative interactions between glycan-binding proteins (GBPs) and enantiomers of common glycans, which were not part of the original training datasets. l -glycans are rare in nature but are important in consideration of safety of putative mirror-image life-forms. Current ML models that predict properties of glycans from their monosaccharide composition cannot extrapolate properties of mirror glycans. Instead, MCNet uses an atom-level description of the glycan to output an estimate of binding to GBPs. MCNet is trained using data from glycan microarrays and affinity measurements unified using a “fraction bound” parameter. Trained MCNet predicted unexpected binding of l -glucose to some fucose-binding GBPs. Both glycan and lectin arrays conformed these predictions. ML models akin to MCNet reach beyond traditional glycobiology and make it possible to anticipate interaction between biomolecules in mirror-life forms and present-day life-forms.

To accelerate the rational design of glycomimetic inhibitors, based on derivatization of a carbohydrate ligand, we introduce a computational pipeline that automates the creation and modeling of analogs and computes their interaction energies. Putative glycomimetics are assembled by grafting small drug-like moieties onto the native carbohydrate scaffold in the presence of the receptor protein, with the moieties chosen from a virtual library of more than 1500 molecular fragments, selected for their synthetic accessibility. The method is illustrated for the case of glycomimetics but is generalizable to any bound ligand. A genetic algorithm (GA) was developed to identify the most likely orientation of the appended moieties in the receptor binding site. For validation, curated experimental data sets were assembled from the literature, consisting of 119 glycomimetics, with reported solution binding free energies, including 46 with corresponding high-resolution crystal structures of the glycomimetic complexes. These data sets were subdivided for protocol testing and "real-world" performance validation. The GA search resulted in an average root-mean-squared deviation (RMSD) of 1.5 Å for the added moieties, compared to their crystallographic data. The GA-generated structures were then subjected to molecular dynamics (MD) simulation, and the performance was evaluated for three post-MD approaches to computing interaction energies: the scoring function from AutoDock Vina-Carb, as well as the generalized Born and Poisson-Boltzmann surface area (GBSA/PBSA) implementations within the AMBER molecular mechanical (MM) force field. For the Test data set of structures with reported energies, the highest coefficient of determination (R² = 0.67) was obtained with MM-PBSA when ligand conformational entropies were included. Current limitations of the protocol and experimental data sets are discussed.

Brucellosis continues to be a significant global zoonotic infection, with diagnosis largely relying on the detection of antibodies against the Brucella O-polysaccharide (O-PS) A and M antigens. In this study, computational methods, including homology modeling, molecular docking, and molecular dynamics simulations, were applied to investigate the interaction of the four murine monoclonal antibodies (mAbs) YsT9.1, YsT9.2, Bm10, and Bm28 with hexasaccharide fragments of the A and M epitopes. Through stringent stability criteria, based on interaction energies and mobility of the antigens, high-affinity binding of A antigen with YsT9.1 antibody and M antigen with Bm10 antibody was predicted. In both the complexes hydrophobic interactions dominate the antigen–antibody binding. These findings align well with experimental epitope mapping, indicating YsT9.1’s preference for internal sequences of the A epitope and Bm10’s preference for internal elements of the M epitope. Interestingly, no stable complexes were identified for YsT9.2 or Bm28 interacting with A or M antigen. This study provides valuable insights into the mechanism of molecular recognition of Brucella O-antigens that can be applied for the development of improved diagnostics, synthetic glycomimetics, and improved vaccine strategies.